You are on page 1of 242

THE HONG KONG POLYTECHNIC UNIVERSITY

DEPARTMENT OF CIVIL AND STRUCTURAL ENGINEERING




THREE DIMENSIONAL SLOPE STABILITY ANALYSIS
AND FAILURE MECHANISM

By

Wen-Bing WEI
BEng, MSc

Supervisor: Dr. Y. M. Cheng

A Thesis Submitted in Partial Fulfilment of the Requirements
for the Degree of Doctor of Philosophy



J uly 2008

Abstract of thesis entitled
THREE DIMENSIONAL SLOPE STABILITY ANALYSIS AND FAILURE
MECHANISM
submitted by Wen-Bing Wei
for the degree of Doctor of Philosophy
at The Hong Kong Polytechnic University in J uly 2008
For slope stability problem, two-dimensional analysis is commonly used for simplicity,
though all slope failures are three-dimensional (3D) in nature. There are only limited
applications of 3D analysis due to the various limitations of three-dimensional slope
stability methods. Recently, there are various important progresses for 3D analysis. 3D
limit equilibrium method (LEM) for general asymmetrical problem together with
innovative optimization method for locating the general critical 3D failure surface has
recently been developed. In addition, 3D strength reduction method (SRM) can now be
conducted within a tolerable duration. Until now, there is still a lack of detailed
investigation of 3D slope stability failure mechanism and the application of 3D LEM
and SRM under different 3D conditions. This study aims to conduct an extensive three-
dimensional slope stability analysis and to investigate the failure mechanism under
different situations. The 3D effect considered in this study includes the important
factors such as slope geometry, boundary conditions, water, soil nail and pile.

Both the LEM and SRM are conducted in this study, and some interesting differences
between these two methods are discovered. It is concluded that both methods have their
own merits and limitations, and a good understanding of these methods are required
before a good solution can be obtained. The suitability of these two methods under
different conditions is investigated and precautions when these methods are applied are
suggested.
In this study, some results which appear to be different from common understanding are
obtained. Some of the results are also different from published studies, and careful
investigations have revealed that the present detailed studies have provided a better and
more reasonable understanding about the failure and stabilization mechanism. For
example, for a simple slope extending to infinity, the critical slip surface is still
basically two-dimensional until the external loading is large to induce a three-
dimensional failure. The failure mechanisms due to the self weight of soil and external
loads are actually different. The discretization domain required for SRM analysis is
found to decrease with the increase of external loads which is also out of expectation.
The distribution of tension force in soil nail is found to be influenced by the state of the
slope (service state, limit state) and the failure modes (external failure, internal failure).
In general, the line of the maximum tension may not correspond to the critical slip
surface as commonly believed, except when the failure mode is an internal tensile
failure. For slope supported with one row of piles, the slip surface is divided into two
parts when the pile spacing is very small, and these two parts gradually become
connected to form a clear single slip surface mechanism with the increase of pile
spacing. The point of maximum shear force in the pile which is commonly used to
determine the location of the slip surface in traditional design is found to be not a valid
assumption, and this is important for design of slopes reinforced with piles. Some
engineers have questioned the disturbance of the seepage pattern due to the presence of
soil nails and piles. Such blocking effect is however found to be negligible.

With the detailed study on 3D failure and stabilization mechanism, a clearer and better
understanding of three-dimensional slope failure has been achieved in the present study.


LIST OF PUBLICATIONS
Wei, W. B., Cheng, Y. M., and Li, L. (2008). Three-dimensional slope failure analysis
by the strength reduction and limit equilibrium methods. Computers and
Geotechnics, In Press.
Wei, W. B. and Cheng, Y. M. (2008). Soil nailed slope by the strength reduction and
limit equilibrium methods, submitted to International J ournal for Numerical and
Analytical Method in Geomechanics.
Wei, W. B. and Cheng, Y. M. (2008). Strength reduction analysis for slope reinforced
with one row of piles, submitted to J ournal of Geotechnical and Geoenvironmental
Engineering.
Wei, W. B. and Cheng, Y. M. (2008). Stability analysis of slope with water flow by
strength reduction method, submitted to Soils and Foundations.
Cheng, Y. M., Lansivaara, T., and Wei, W. B. (2008). Reply to Comments on Two-
dimensional slope stability analysis by limit equilibrium and strength reduction
methods. Computers and Geotechnics, Vol. 35, No.2, 309-311
Cheng, Y. M., Lansivaara, T., and Wei, W. B. (2007). Two-dimensional slope stability
analysis by limit equilibrium and strength reduction methods. Computers and
Geotechnics, Vol. 34, No.3, 137-150
Wei, W. B. and Cheng, Y. M. (2007). Strength reduction method for three dimensional
slope stability analysis. 60
th
Canadian Geotechnical Conference and 8
th
J oint
CGS/IAH-CNC Groundwater Conference (Ottawa, Ontario, Canada, October 21-
24, 2007), 815-820
Cheng, Y. M., Wei, W. B., and Lansivaara T. (2006). Factors of safety by limit
equilibrium and strength reduction methods. Numerical Methods in Geotechnical
Engineering Schweiger (ed.), Sixth European Conference on Numerical Methods
in Geotechnical Engineering (Graz, Austria, 6-8 September 2006), 485-490
ACKNOWLEDGEMENTS
I would like to express my great gratitude and sincere thanks to my supervisor, Dr. Y.
M. Cheng, for his continued inspiration, support and invaluable guidance during my
study. Without his endless efforts and experienced guidance, it would not have been
possible to complete this work. I also appreciate his caring about my life.

Financial support for this research work from The Hong Kong Polytechnic University is
greatly acknowledged. Also, administrative and technical supports from the staff of the
Department of Civil and Structural Engineering are deeply appreciated.

I am very thankful to Professor J . A. Wang of University of Science and Technology
Beijing who introduced me to the world of geotechnical engineering. Many thanks are
due to Professor H. G. J i and Professor Z. H. Chen of University of Science and
Technology Beijing for their useful suggestions and discussions. Thanks are also
extended to my friends Dr. Z. Fang, Dr. L. J . Su, Dr. W. H. Zhou and Dr. H. H. Zhu for
their fruitful discussions.

I would also like to express my sincere gratitude to the members of my oral examination
committee, Professor M. W. Xie, Dr. J . Yang and Professor J . H. Yin for reviewing the
manuscript and offering valuable comments and suggestions.

I deeply appreciate my parents and my brother for their constant support and endless
encouragement. Finally, I wish to express my special thanks to my wife, Lan-J uan Xiao,
for her support and understanding.

TABLE OF CONTENTS
CERTIFICATE OF ORIGINALITY
ABSTRACT
LIST OF PUBLICATIONS
ACKNOWLEDGEMENTS
TABLE OF CONTENTS
LIST OF TABLES
LIST OF FIGURES
CHAPTER 1: INTRODUCTION 1
1.1 Background and motivation 1
1.2 Objective 5
1.3 Organization of thesis 5
CHAPTER 2: LITERATURE REVIEW 8
2.1 Slope stability analysis methods 8
2.1.1 Limit equilibrium method 8
2.1.2 Variational calculus method 11
2.1.3 Limit analysis method 12
2.1.4 Strength reduction method 14
2.2 Stability analysis of soil nailed slope 18
2.2.1 Soil nail resistance 19
2.2.2 Failure modes of soil nailed slope 21
2.2.3 Design methods of soil nailed slope 22
2.2.4 Full scale and small scale test of soil nailing slope 25
2.3 Stability analysis of slopes reinforced with piles 27
I
2.4 Stability analysis of slopes with water flow 30
CHAPTER 3: TWO-DIMENSIONAL SLOPE STABILITY ANALYSIS BY
LIMIT EQUILIBRIUM AND STRENGTH REDUCTION
METHODS 34
3.1 Introduction 34
3.2 Stability analysis for a simple and homogeneous soil slope 36
3.3 Stability analysis of a slope with a soft band 39
3.4 Local minimum in LEM 43
3.5 Influence of elastic modulus for SRM analysis 45
3.6 Discussion and Conclusion 45
CHAPTER 4: THREE-DIMENSIONAL SLOPE FAILURE ANALYSIS BY THE
STRENGTH REDUCTION AND LIMIT EQUILIBRIUM
METHODS 65
4.1 Introduction 65
4.2 Stability analysis for a vertical cut 65
4.3 Stability analysis for a vertical cut with a weak layer 70
4.4 Stability analysis for a slope with transverse earthquake load 72
4.5 Failure mode due to self weight for a simple infinite slope 73
4.6 Stability analysis for a locally loaded slope 74
4.7 Curvature effect on the slope stability 78
4.8 Discussion and conclusion 81
CHAPTER 5: SOIL NAILED SLOPE BY STRENGTH REDUCTION AND
LIMIT EQUILIBRIUM METHODS 98
5.1 Introduction 98
II
5.2 Importance of nail head in analysis 101
5.3 Soil nailed slope stability with different soil properties 103
5.4 Slope with different nail inclination angle 105
5.5 Slope with different nail length 106
5.6 Slope with different soil nail layout 107
5.7 Soil nailed slope with external pressure on the top 108
5.8 Influence of nail elastic modulus 109
5.9 Analysis of a vertical soil nailed wall in Seattle, Washington 110
5.10 Distribution of the nail tension force and critical slip surface 111
5.11 Slip surface for face failure 116
5.12 Influence of nail spacing on the failure modes 117
5.13 Discussion and Conclusion 117
CHAPTER 6: EFFECTS OF CURVATURE AND LOCAL LOADING ON 3D
SOIL NAILED SLOPE 142
6.1 Introduction 142
6.2 Numerical simulation for locally loaded soil nailing slope 142
6.3 Numerical simulation for idealized convex and concave slopes 144
6.4 Numerical simulation for intersected slopes 145
6.5 Some small scale model testing results 147
6.6 Discussion and Conclusion 148
CHAPTER 7: STRENGTH REDUCTION ANALYSIS FOR SLOPE REINFORCED
WITH ONE ROW OF PILES 159
7.1 Introduction 159
7.2 Failure mode of slopes with different pile spacing 160
III
7.3 Upper and lower bound of the factor of safety 162
7.4 Optimization of the pile position 163
7.5 Conclusion 166
CHAPTER 8: STABILITY ANALYSIS OF SLOPE WITH WATER FLOW BY
STRENGTH REDUCTION METHOD 177
8.1 Introduction 177
8.2 Stability analysis for a simple slope with seepage flow 178
8.3 Stability analysis for a simple slope with irregular pore pressure 180
8.4 Stability analysis for soil nailing slope with seepage flow 180
8.5 Stability analysis for piled slope with seepage flow 182
8.6 Stability analysis for locally loaded slope with seepage flow 183
8.7 Conclusion 184
CHAPTER 9: CONCLUSIONS AND RECOMMENDATIONS 193
9.1 Conclusions 193
9.1.1 Comments on limit equilibriumand strength reduction methods 193
9.1.2 Main findings from analysis of non-reinforced slopes with obvious
3D effects 194
9.1.3 Main findings from analysis of soil nailing slope 195
9.1.4 Main findings from analysis of piled slope 196
9.1.5 Main findings from analysis of slopes with water flow 197
9.2 Recommendations and suggestions 198
REFERENCES 200


IV
LIST OF TABLES
Table 3.1 Factors of safety by LEM and SRM 50
Table 3.2 Soil properties for Figure 3.6 51
Table 3.3a FOS by SRM from different programs when c=0 and =25 for soft band.
The values in each cell are based on SRM1 and SRM2 respectively. (min.
FOS=0.927 from Morgenstern-Price analysis) 51
Table 3.3b FOS by SRM from different programs when =0 and c=10 kPa for soft
band. The values in each cell are based on SRM1 and SRM2 respectively.
(min. FOS=1.03 from the Morgenstern-Price analysis) 51
Table 3.4 FOS with non-associated flow rule for 12m domain 52
Table 3.5 FOS with associated flow rule for 12m domain 52
Table 3.6 Comparison of factor of safety for Figure 3.14 53
Table 3.7 Comparison of factor of safety for Figure 3.15 53
Table 3.8 Comparison of factor of safety for Figure 3.16 53
Table 3.9 Comparison of factor of safety for Figure 3.17 53
Table 4.1 The dependence of FOS on convergence criterion and iteration number 84
Table 4.2 Comparison of SRM and LEM results with various earthquake loads in the
x-direction 84
Table 4.3 Comparison of SRM and LEM results with various earthquake loads in the
y-direction 84
Table 4.4 Safety factors under the local loading of 100kPa 84
Table 4.5 Safety factors with different model lengths, obtained by the SRM 85
Table 4.6 Variation of FOS respect to curvature (no load) 85
Table 4.7 Variation of FOS respect to curvature (with 4m width load 100kPa) 85
V
Table 5.1 Parameters of grout-soil-nail system 122
Table 5.2 Factor of safety for nail head with different elastic modulus 122
Table 5.3 Factors of safety by LEM and SRM 122
Table 5.4 Results for different nail inclination 123
Table 5.5 Factor of safety for different soil nail length 123
Table 5.6 Factor of safety with 200 kPa top pressure (bond load controlled by
overburden stress) 123
Table 5.7 Factor of safety with 200 kPa top pressure (constant pull out resistance) 124
Table 5.8 Factor of safety with different nail elastic modulus (slope angle 45 degree) 124
Table 5.9 Factor of safety with different nail elastic modulus (vertical cut slope) 124
Table 6.1 Variation of FOS respect to curvature (nail 10m, no load) 150
Table 6.2 Variation of FOS respect to curvature (nail 10m, with 4m width load 100kPa) 150
Table 7.1 Factor of safety with different pile spacing 168
Table 8.1 Factor of safety for different situations by SRM 186
Table 8.2 Factor of safety of nailed slope by SRM 186










VI
LIST OF FIGURES
Figure 2.1 Failure modes for reinforced soil slopes (after Elias et al., 2001) 33
Figure 2.2 Potential soil nail wall internal failure modes (after Byrne et al., 1998) 33
Figure 3.1 Discretization of a simple slope model 54
Figure 3.2 Slip surface comparison with increasing friction angle (c=2kPa) 54
Figure 3.3 Slip surface comparison with increasing cohesion (=5) 55
Figure 3.4 Slip surface comparison with increasing cohesion (=35)55
Figure 3.5 Slip surface comparison with increasing cohesion (=0) 56
Figure 3.6 A slope with a thin soft band 56
Figure 3.7 Mesh plot of the three numerical models with a soft band 57
Figure 3.8 Critical failure surfaces from LEM and SRM for frictional soft band problem58
Figure 3.9 Critical solutions from LEM and SRM when the bottom soil layer is weak 59
Figure 3.10 Slope geometry and soil property under study 59
Figure 3.11 Result derived by SRM 60
Figure 3.12 Global and local minima by LEM 60
Figure 3.13 Local minima from LEM and critical solution from SRM for a simple slope 62
Figure 3.14 Slope geometry with two soils and a horizontal boundary 63
Figure 3.15 Slope geometry with two soils and a vertical boundary 63
Figure 3.16 Slope geometry with two soils and an inclined boundary 63
Figure 3.17 Slope geometry with three soils and an inclined boundary 64
Figure 3.18 Critical slip surface of case 2 in Figure 3.17 based on SRM and LEM 64
Figure 4.1 Slip surface of a vertical cut with two unconstrained vertical planes
(FOS=1.51) 86
Figure 4.2 Slip surface of a vertical slope with only one unconstrained vertical plane
(FOS=1.55) 86
VII
Figure 4.3 The geometry of the three-dimensional wedge block 87
Figure 4.4 The geometry of the two-dimensional wedge block 87
Figure 4.5 Variation of FOS with respect to the dip angle of the sliding plane 87
Figure 4.6 A large model for a vertical cut (FOS=1.56) 88
Figure 4.7 Displacement versus factor of safety 88
Figure 4.8 Non-linear analysis results using different strength reduction factors 89
Figure 4.9 Geometry of a vertical cut with an inclined weak layer (after Huang and Tsai,
2000) 89
Figure 4.10 Model one of the strength reduction analysis for a vertical cut with a weak
layer 90
Figure 4.11 Model two of the strength reduction analysis for a vertical cut with a weak
layer 91
Figure 4.12 Model three of the strength reduction analysis for a vertical cut with a weak
layer 92
Figure 4.13 Mesh for a slope with transverse earthquake load 92
Figure 4.14 Simple slope stability analysis with varying heterogeneity 93
Figure 4.15 The geometry of the slope under local loading 93
Figure 4.16 The slip surfaces for different loading lengths when B=2m 93
Figure 4.17 The slip surfaces for different model lengths when L/B=4 94
Figure 4.18 The slip surfaces for different local loadings when L/B=4 and model
length=20m 94
Figure 4.19 The geometry of the slope section 94
Figure 4.20 Typical geometry of convex and concave slope 95
Figure 4.21 Mesh plot and slip surface of concave and convex models 95
Figure 4.22 Slip surfaces at different curvatures (no load) 95
VIII
Figure 4.23 Slip surfaces for different curvatures (no load) 96
Figure 4.24 Slip surfaces for different curvatures with 200kPa loading 96
Figure 4.25 Vertical cut slope with 3m width 50kPa loading along the edge of the crest 96
Figure 4.26 Vertical cut slope with 3m long and 3m wide 50kPa loading 97
Figure 5.1 Idealization of grout-cable system (from Itasca, 2006) 125
Figure 5.2 Idealization of soil nail system 125
Figure 5.3 Plot of the soil nailed slope model 126
Figure 5.4 Slip surface and the tension stress of soil nail without nail head (FOS=1.20) 126
Figure 5.5 Slip surface for soil nailed slope without nail head by SRM and compared
with LEM results 127
Figure 5.6 Slip surface and the tension stress of soil nail for model with nail head
(FOS=1.28) 127
Figure 5.7 Slip surface for soil nailed slope by SRM for the model with nail head 127
Figure 5.8 Slip surface of the slope with 0.14m width and 0.15m height nail head 128
Figure 5.9 Slip surface comparison with increasing friction angle (c=2kPa) 128
Figure 5.10 Slip surface comparison with increasing friction angle (c=5kPa) 129
Figure 5.11 Slip surface comparison with increasing friction angle (c=20kPa) 129
Figure 5.12 Slip surface obtained by different element size (c=5kPa, =35) 130
Figure 5.13 Slip surface and nail axial force distribution for different nail inclined angle
(nail simulated by cable element) 130
Figure 5.14 Slip surface and nail axial force distribution for different nail inclination
(bending effect considered) 131
Figure 5.15 Slip surface and nail bending moment distribution for different nail
inclination (bending effect considered) 131
Figure 5.16 Slip surface and the nail tension force for different nail length 132
IX
Figure 5.17 Slip surface and the tension stress in different layout with zero inclined
angle 132
Figure 5.18 Slip surface and the nail load for different layout with 20 nail inclination 133
Figure 5.19 Results for different layout with zero inclined angle (constant nail pullout
strength) 133
Figure 5.20 Results for different layout with 20 nail inclination (constant nail pullout
strength) 133
Figure 5.21 Vertical cut soil nailed slope model 134
Figure 5.22 Slip surface and nail tension stress of the vertical soil nailed wall with
FOS=1.76 134
Figure 5.23 Slip surface of the vertical soil nailed wall by LEM with FOS=1.90 134
Figure 5.24 Slip surface of the vertical soil nailed wall with high shear strength at left
part by SRM (FOS=2.10) 135
Figure 5.25 Slip surface of an excavated vertical cut with applied pressure 200kPa at
left corner by SRM (FOS=2.29) 135
Figure 5.26 Load transfer mechanism in soil nails (after Byrne et al., 1998) 135
Figure 5.27 Axial tensile force distribution of soil nail in different external failure modes 136
Figure 5.28 Slip surface and nail tension stress distribution in different internal failure
modes 136
Figure 5.29 Axial tensile force distribution of soil nail in different state 136
Figure 5.30 Axial tensile stress distribution in different state for a vertical cut slope 137
Figure 5.31 Axial tensile stress distribution in different limit state 137
Figure 5.32 Nail force distribution in limit state 138
Figure 5.33 Nail force distribution in service state 139
Figure 5.34 Slip surface of soil 1 with face failure in different nail interval 140
X
Figure 5.35 Slip surface of soil 2 with face failure in different nail interval 140
Figure 5.36 Slip surface and nail force for slope with 10m horizontal nail interval and
soil-nail interface strength is half of the soil strength (FOS=1.25) 141
Figure 5.37 Slip surface and nail force for slope with 10m horizontal nail interval and
soil-nail interface strength is the same as the soil strength (FOS=1.30) 141
Figure 6.1 The geometry of the slope under local loading 151
Figure 6.2 The slip surface for different loading length when B=2m (nail 8m,
load=200kPa) 151
Figure 6.3 Nail maximum tension force distribution for L/B=1 151
Figure 6.4 Nail maximum tension force distribution for L/B=2 152
Figure 6.5 Nail maximum tension force distribution for L/B=4 152
Figure 6.6 Nail maximum tension force distribution with no loading (FOS=1.94) 152
Figure 6.7 The geometry of the slope section 153
Figure 6.8 Mesh plot of concave and convex models reinforced with nails 153
Figure 6.9 Slip surfaces at different curvatures (with nail, no load) 153
Figure 6.10 Slip surfaces for different curvatures with 200kPa loading (no load, nail 8m) 154
Figure 6.11 Slip surfaces for different curvatures with 200kPa loading (nail 8m, pullout
failure) 154
Figure 6.12 Slip surfaces for different curvatures with 200kPa loading (nail 8m, face
failure) 154
Figure 6.13 Nail maximum tension force distribution for a 135 concave slope with no
loading 155
Figure 6.14 Nail maximum tension force distribution for a 145 convex slope with no
loading 155
Figure 6.15 Nail maximum tension force distribution for a 135 concave slope with
XI
200kPa loading 155
Figure 6.16 Nail maximum tension force distribution for a 145 convex slope with
200kPa loading 156
Figure 6.17 Nail maximum tension force distribution for a 135 concave slope with
200kPa loading (face failure) 156
Figure 6.18 Nail maximum tension force distribution for a 145 convex slope with
200kpa loading (face failure) 156
Figure 6.19 Nail maximum tension force distribution for a no curvature slope with
200kpa loading (face failure) 157
Figure 6.20 Slip surface of vertical cut slope with 3m long and 3m wide 50kPa loading
(pullout failure) 157
Figure 6.21 Slip surface of vertical cut slope with 3m long and 3m wide 50kPa loading
(face failure) 158
Figure 6.22 Different failure shapes after failure mass was removed (after Tsui, 2007) 158
Figure 7.1 Slope model and finite difference mesh 169
Figure 7.2 Slip surface of the slope with no pile (FOS=1.20) 169
Figure 7.3 Slip surface at different sections for s=2D (FOS=1.78) 169
Figure 7.4 Slip surface at different sections for s=3D (FOS=1.72) 170
Figure 7.5 Slip surface at different sections for s=4D (FOS=1.61) 170
Figure 7.6 Slip surface at different sections for s=5D (FOS=1.55) 171
Figure 7.7 Slip surface at different sections for s=6D (FOS=1.52) 171
Figure 7.8 Slip surface at different sections for s=8D (FOS=1.42) 172
Figure 7.9 Slip surface at the section of soil midway between piles 172
Figure 7.10 Slip surface obtained by SRM based on extreme point of shear force and comparison
with critical slip circle obtained by Bishops simplified method 173
XII
Figure 7.11 Shear force distribution of the piled slope for s=3D 173
Figure 7.12 Slip surface and mesh for the slope with pile wall (FOS=1.89) 173
Figure 7.13 Factor of safety with respect to different pile spacing 174
Figure 7.14 Slip surface and mesh for the slope with strong soil at upper part
(FOS=1.94) 174
Figure 7.15 Slip surface and mesh for the slope with strong soil at lower part
(FOS=1.87) 174
Figure 7.16 Slip surface for the slope with pile wall installed at 0.2m towards the slope
crest as measured from the middle of slope 175
Figure 7.17 Pile position for the slope with soil cohesion 10kpa and friction angle 20
degree 175
Figure 7.18 Slip surface of the slope with no pile for soil cohesion=20kpa and friction
angle 10 degree (FOS=1.14) 175
Figure 7.19 Slip surface for the slope with pile installed in the middle of slope (soil
cohesion=20kpa, friction angle=10 degree) 176
Figure 7.20 Slip surface for the slope with pile installed 0.55m upper of the middle of
slope (soil cohesion=20kpa, friction angle=10 degree) 176
Figure 7.21 Pile position for the slope with soil cohesion 20kpa and friction angle 10
degree 176
Figure 8.1 Pore water pressure and flow vector of a simple slope from a free-surface
seepage analysis 187
Figure 8.2 Slip surface for slope with cohesion 1kPa and friction angle 45 187
Figure 8.3 Slip surface for slope with cohesion 2kPa and friction angle 45 187
Figure 8.4 Slip surface for slope with cohesion 5kPa and friction angle 35 188
Figure 8.5 Slip surface for slope with cohesion 10kPa and friction angle 25 188
XIII
Figure 8.6 Water pressure by water table (piezometric line) assuming hydrostatic
condition 188
Figure 8.7 Slip surface and pore pressure with water block at upper and bottom left for
slope with cohesion 2kPa and friction angle 45 189
Figure 8.8 Slip surface and pore pressure with water block wall for slope with cohesion
2kPa and friction angle 45 189
Figure 8.9 Model plot of the soil nailing slope 189
Figure 8.10 Pore pressure distribution of the soil nailing slope with water blocking
effect 190
Figure 8.11 Slip surface for the nailed slope with water flow (FOS=0.95) 190
Figure 8.12 Model plot of the piled slope 190
Figure 8.13 Pore pressure and slip surface of the slope without pile (FOS=0.85) 191
Figure 8.14 Pore pressure distribution of the piled slope with water blocking effect 191
Figure 8.15 Slip surface for the piled slope with water flow (FOS=1.29) 191
Figure 8.16 Slip surface for the slope with pile installed at 2.0m towards the slope toe as
measured from the middle of slope (FOS=1.34) 192
Figure 8.17 Pore water pressure and slip surface for the locally loaded slope with water 192

XIV
Chapter 1: Introduction

- 1 -

CHAPTER 1: INTRODUCTION

1.1 Background and motivation
Due to the rapid increase in population and huge demand for infra-structures, many
buildings and highways are constructed adjacent to natural slopes and many cut slopes
are formed for various purposes in Hong Kong and other developed cities. Landslides
have been a major cause of disasters resulting in considerable loss of human lives and
property damages in hilly terrain. Slope stability problem is, therefore, one of the most
commonly occurred geotechnical problems in Hong Kong and many other developed
cities. Usually, slope failure occurs as a result of triggering mechanisms, in which water
is a main factor for slope geo-disasters. The infiltration of water into soil increases pore
water pressure and decreases the shear strength of the soil structure, which will then
leads to slope instability.

In order to increase the safety factor to the required level, the use of reinforcement is
now commonly adopted for slope stabilization. For ease of construction and some other
potential benefits, soil nailing has been proven to be a practical technique for
stabilization of slopes (Bruce and J ewell, 1986). Passive soil nailing is one of the most
widely used slope stabilization measures in Hong Kong over the last two decades. It is
currently the predominant method for upgrading the stability of existing soil cut slopes.
Installing piles into the slope to improve the stability has also been successful and is
also demonstrated to be an effective approach (Ito and Matsui, 1975; Reese et al., 1992).

Chapter 1: Introduction

- 2 -
For slope stability analysis, two-dimensional (2D) plane strain approach is usually
employed for simplicity, and the most extensively used method is the limit equilibrium
method (LEM) which is well known to be a statically indeterminate problem, and
assumptions on the internal force distribution are required for the solution of the factor
of safety (Bishop, 1955; Morgenstern and Price, 1965; Spencer, 1967; J anbu, 1973).
The calculus of variation approach by Baker and Garber (1978) does not require the
assumption on the internal force distribution, but it is not easy to be used in practice.
Besides, limit analysis has also been used for simple problems (Chen 1975), but its
applications in complicated real problems are still limited, and this method is seldom
adopted for routine analyses and designs. In recent years, there is a great development
of the shear strength reduction method (SRM) for slope stability analysis (Zienkiewicz
et al., 1975; Ugai and Leshchinsky, 1995; Dawson et al., 1999; Griffiths and Lane,
1999). This technique is also adopted in several well-known commercial geotechnical
finite element programs. The SRM does not require inter-slice force assumption, and
the main advantage is that it can automatically determine the slip surface and is suitable
to many complicated conditions.

Two-dimensional analyses, though helpful for designing most of the slopes, are not
applicable to many situations in which three-dimensional (3D) effect is obvious. The
common approaches for three-dimensional slope stability analysis are still the LEM,
which are usually the direct extensions of the various two-dimensional methods. Most
of these methods are based on the assumption that the failure mass is symmetrical with
a known sliding direction, so a true asymmetric slope failure cannot be modelled
directly by the classical three-dimensional methods. The lack of suitable method for
locating the critical general 3D failure surface is another major limitation of the 3D
LEM. There are also several three-dimensional limit analysis models (Giger and Krizek,
Chapter 1: Introduction

- 3 -
1975; Michalowski, 1989; Chen et al., 2001a; Farzaneh and Askari, 2003) in literature.
The construction of the three-dimensional failure mechanism for limit analysis is
difficult for a complicated slope, and this approach is seldom adopted except for simple
slopes. By contrast, the SRM appears to be simple for extension to three-dimensional
analysis, since this method can be employed within a single framework for both two-
and three-dimensional slopes. Some special limitations of the SRM are discovered in
this study and will be illustrated later.

All slope failures are three-dimensional in nature, especially for some natural slopes or
slopes with transverse loads or concentrated loads. There is however only limited
application of 3D analysis due to the various limitations of the three-dimensional slope
stability methods mentioned previously. Recently, there are various important
progresses for 3D analysis. Huang and Tsai (2000) have proposed the first method for
3D asymmetrical slope stability analysis where the sliding direction enters into the
direct determination of the factor of safety (FOS), but the sliding direction will be
different for different soil columns. Cheng and Yip (2007) have developed a new
asymmetric 3D analysis model under which there is only one sliding direction for the
whole failure mass, and this assumption has overcome the convergence problem under
transverse load in the Huang and Tsai formulation (2000). Cheng et al. (2005) have
proposed the use of 3D NURBS surface and the simulated annealing method to locate
the critical general failure surface. In addition, with the development of computer
hardware and software, 3D SRM can now be conducted within a tolerable time span and
may become a prospective approach for complicated 3D slope analysis.

Due to the improvement of 3D LEM and 3D SRM mentioned above, it can be
anticipated that the three-dimensional slope stability analysis will be more popular in
Chapter 1: Introduction

- 4 -
the future if the techniques are mature. Up to present, there is still a lack of detailed
investigation of 3D methods and their application under different 3D conditions.

In view of the above, this study aims to conduct an extensive three-dimensional slope
stability analysis and to investigate the failure mechanism under different situations.
The 3D effect considered in this study lies on two aspects. Firstly, for slopes with no
reinforcement, the 3D analysis should be conducted in the following situations: the
slope geometry is irregular, the soil properties are not homogeneous, the slope is subject
to concentrated load, or the boundary conditions cannot be ignored. Secondly, for slopes
with reinforcement, even though the slope geometry and boundary conditions have no
obvious 3D effect, if the reinforcement is not continuous along the slope, 3D analysis
should also be conducted. For example, if the reinforcement is geotextile, it is
reasonable to conduct 2D analysis. If the reinforcement is soil nail or pile, 3D analysis
should be preferred for better modelling of the nail/pile-soil interaction.

In this study, both the limit equilibrium and strength reduction methods are used. The
strength reduction analysis is conducted for all the cases involved in this thesis, since
SRM is powerful in automatically detecting the failure surface without any optimization
search technique which is usually required in the LEM analysis, and it is also powerful
in modelling the interaction between soil and reinforcement. Some of the results are
also compared with the limit equilibrium analysis, and these useful comparisons have
revealed the merits and limitations of the LEM and SRM which are important for the
proper analysis of slope stability problem. Slopes with and without reinforcement are
both considered, and useful and important results are obtained. Besides, the slope with
water flow is also investigated. Some interesting and useful findings are obtained from
Chapter 1: Introduction

- 5 -
this analysis, and different failure modes are found for slopes with and without
reinforcement.

1.2 Objective
The main objectives of this research are as follows:

(a) Study the stability and failure mechanism of non-reinforced slopes with obvious 3D
effects in geometry or boundary conditions, such as locally loaded slopes.

(b) Study the stability and failure mechanism of soil nailed slopes without obvious 3D
effects in slope geometry.

(c) Study the stability and failure mechanism of soil nailed slopes with obvious 3D
effects in slope geometry.

(d) Study the stability and failure mechanism of pile supported slopes.

(e) Study the stability and failure mechanism of slopes with water flow.

(f) Investigate the suitability of 3D LEM and 3D SRM approaches under different
conditions, and provide suggestions and precautions in employing these methods.

1.3 Organization of thesis
This thesis consists of nine chapters which are organised as follows:

Chapter 1: Introduction

- 6 -
Chapter 1 is the current chapter which briefly introduces the background and objectives
of this research.

Chapter 2 reviews previous investigations concerning the present research in four
aspects: 2D and 3D slope stability analysis methods which includes limit equilibrium
methods, variational calculus methods, limit analysis methods, strength reduction
methods; stability analysis of soil nailing slope; stability analysis of slopes reinforced
with piles; stability analysis of slopes with water flow.

Chapter 3 gives an extensive comparison on the factors of safety and the locations of
critical failure surfaces obtained by the limit equilibrium method and strength reduction
method for various 2D slopes. The advantages and limitations of the strength reduction
and limit equilibrium methods are illustrated and some useful and surprising results are
also found.

Chapter 4 analyzes several typical non-reinforced slopes with obvious 3D effects by
3D strength reduction method and 3D limit equilibrium method. The suitability of these
two methods under different conditions is investigated and precautions for these
methods are suggested. Some surprising results are also found and are explained.

Chapter 5 presents a detailed strength reduction study of nailed slope under different
conditions. A more realistic 3D analysis is carried out to consider the interaction
between soil and nail, while the slope geometry has no apparent 3D effect. The results
are also compared with those obtain by the LEM.

Chapter 1: Introduction

- 7 -
Chapter 6 is devoted to investigation of some soil nailed slopes with obvious 3D
effects in geometry and boundary conditions (locally loaded slope, idealized curvature
slope, and intersected slope).

Chapter 7 conducts 3D SRM analysis for the slope reinforced with one row of piles
and the failure mode is determined with respect to different pile spacing. The upper and
lower bounds of the factor of safety, and the optimal pile location are also discussed.

Chapter 8 employs the strength reduction method on slope stability analysis with water
flow. Several 2D and 3D slope models with and without reinforcement are investigated
and discussed.

Chapter 9 summarizes the major findings from this research and suggests some areas
which require further study.




Chapter 2: Literature Review

- 8 -

CHAPTER 2: LITERATURE REVIEW

The review of previous investigations relevant to the present research is described in
this chapter, and it is mainly divided into four parts. The first part is an introduction to
the slope stability analysis methods. The second part presents the stability analysis of
soil nailed slope. The third part is devoted to the stability analysis of slopes reinforced
with piles. The fourth part presents the stability analysis of slopes with water flow.

2.1 Slope stability analysis methods
Slope stability can be analyzed by a number of methods which include limit
equilibrium method, variational calculus method, limit analysis method, strength
reduction method, rigid element method and others. For slope stability analysis, two-
dimensional (2D) plane strain analysis is commonly used for simplicity. All slope
failures are however three-dimensional (3D) in nature, particularly for non-
homogeneous slopes with transverse loads or concentrated loads. In this section, both
the 2D and 3D slope stability methods are briefly summarized.

2.1.1 Limit equilibrium method
For two-dimensional slope stability analysis, the limit equilibrium method (LEM) is the
most widely used approach (Bishop, 1955; Lowe and Karafiath, 1960; Morgenstern and
Price, 1965; Spencer, 1967; J anbu, 1973). The various 2D limit equilibrium methods are
well reviewed and summarized by Fredlund and Krahn (1984), Morgenstern (1992),
Duncan (1996), and Cheng and Zhu (2004). Most of the limit equilibrium methods are
based on the techniques of slices, and assumptions on the interslice forces distribution
Chapter 2: Literature Review

- 9 -
are required for the solution of the factor of safety (FOS). In the traditional LEM
analysis, the assumptions are usually focused on the inclination or the location of the
interslice forces. There is another type of assumption which is focused on the
distribution of the base normal forces on the slip surface, and this approach has been
adopted by Bell (1968), Leshchinsky and Huang (1992a, 1992b). The assumption on the
base normal stress distribution along the slip surface has received little attention for a
long time, but it is reconsidered by Zhu and Lee (2002) to arrive at three equilibrium
equations. Incorporation of these equations can lead to a single cubic function with
respect to the factor of safety, and this function can be solved explicitly. Subsequently,
Zhu et al. (2003) proposed a generalized framework including nearly all of the existing
limit equilibrium methods of slices with general failure surfaces. Different assumptions
about the interslice forces can be converted into a unified form of expression of the
normal stress distribution along the failure surface.

The development of various three-dimensional slope stability methods are basically the
extensions of the corresponding 2D analysis. Cavounidis (1987) has demonstrated that
the factor of safety of a 3D slope should normally be greater than that for a
corresponding two-dimensional slope. 3D problem has one major factor not found in the
corresponding 2D analysis: sliding direction. The common 3D methods include those
by Baligh and Azzouz (1975), Hovland (1977), Chen and Chameau (1982), Azzouz and
Baligh (1983), Hungr (1987), Gens et al. (1988), Zhang (1988), Ugai (1988), Lam and
Fredlund (1993), Huang and Tsai (2000, 2002), Chang (2002), Chen et al. (2003a),
Loehr et al. (2004), and Cheng and Yip (2007). Most of these methods are based on the
assumption that the failure mass is symmetrical with a known sliding direction (usually
zero sliding direction), so an asymmetric slope failure cannot be modeled directly by the
classical three-dimensional methods. J iang and Yamagami (1999) have proposed the
Chapter 2: Literature Review

- 10 -
concept of axes rotation and minimum factor of safety to determine the sliding direction,
but this approach is time consuming in the geometry calculation. Huang and Tsai (2000)
have proposed the first method for 3D asymmetrical slope stability analysis where the
sliding direction enters into the direct determination of the FOS, but the sliding
direction will be different for different soil columns. Cheng and Yip (2007) have
developed a new asymmetric 3D analysis model under which there is only one sliding
direction for the whole failure mass, and this simplification has overcome the
convergence problem under transverse load in the Huang and Tsai formulation (2000,
2002). In this method, the critical general 3D failure surface is located by 3D NURBS
surface and the simulated annealing method proposed by Cheng et al. (2005). In recent
years, the 3D limit equilibrium methods are incorporated with the geographic
information system (GIS) spatial analysis function (Xie et al., 2003, 2004a, 2004b,
2006a, 2006b). This formulation can make use of the spatial functions of GIS for
processing complex spatial data, and render the 3D slope stability problem easier to be
analyzed.

Most of the existing 3D limit equilibrium methods seldom satisfy all the six equilibrium
conditions. By introducing the normal force assumption, which is successfully used in
2D slope limit equilibrium analysis (Zhu and Lee, 2002; Zhu et al., 2003), Zhu and Qian
(2007) have derived a rigorous 3D limit equilibrium solution which satisfies six
equilibrium conditions. The normal stress distribution over the 3D slip surface is
specified by a function with five parameters. The six equilibrium equations are then
reduced to a sixth-order algebraic equation in terms of the 3D factor of safety. A similar
rigorous limit equilibrium method for the three-dimensional stability analysis of slope is
also proposed by Zheng (2007) based on the adjustment of the normal stresses on the
slip surface. The distinct advantage of these two rigorous methods is that it does not
Chapter 2: Literature Review

- 11 -
require the complicated inter-column force assumptions. In order to render the solution
statically determinate, five parameters are introduced to represent the base normal
pressure on the slip surface, and whether the general 3D normal stress distribution for
different complicated conditions can be perfectly described or adjusted by only five
parameters may need some further study.

2.1.2 Variational calculus method
The variational calculus method was first used in 2D slope stability analysis by Baker
and Garber (1978). This approach was subsequently employed by J ong (1980) for
vertical cut analysis in cohesive frictionless soil. This method does not require the
assumption on the internal force distribution, but it is not easy to be used in practical
analysis. Cheng et al. (2008) have developed the numerical algorithm based on the
extremum principle by Pan, and the formulation which relies on the use of modern
heuristic optimization method can be viewed as an equivalent form of the variational
method in a discretized form but is applicable for complicated real problem.

The variational calculus approach has been employed in 3D slope stability analysis by
Leshchinsky et al. (1985), Ugai (1985), Leshchinsky and Baker (1986), Baker and
Leshchinsky (1987), and Leshchinsky and Huang (1992b). In such approaches, the
minimum factor of safety and the associated failure surface can be obtained at the same
time. However, these methods are limited to homogeneous and symmetrical problems,
and further study is required on the application in practical problems with complicated
geometric and loading conditions.



Chapter 2: Literature Review

- 12 -
2.1.3 Limit analysis method
The limit analysis method includes upper bound approach and lower bound approach,
and the general analysis process is the construction of a statically admissible stress field
for the lower bound analysis or a kinematically admissible velocity field for an upper
bound analysis. Optimization analysis of the objective function will then be conducted.
The lower bound approach has been used in 2D slope stability analysis by Chen (1975),
Bottero et al. (1980), Zhang (1999), Kim et al. (2002), and Loukidis et al. (2003), while
the application of this approach in 3D slope stability analysis has been conducted by
Lyamin (1999), Lyamin and Sloan (2002a). In most of these methods, the finite element
discretization and linear programming technique are usually used to obtain the lower
bound solution. Due to the inability of the linear programming technique for
complicated problems with multi local minima, some researchers are using various
advanced optimization method in order to obtain a better solution. Stress fields
employed in the lower bound solutions are usually assumed without an apparent relation
to the actual stress fields, and it is usually not easy to obtain the lower bound solutions
for a practical slope problem. The lower bound approach is hence seldom adopted as
compared with the upper bound approach in slope stability analysis.

The upper bound approach was first used in 2D slope stability analysis by Drucker and
Prager (1952) to determine the critical height of a slope. Subsequently, Chen and Giger
(1971), Chen (1975), Karal (1977a, 1977b), and Izbicki (1981) also applied and
extended the upper bound approaches in 2D slope analysis. Michalowski (1995)
proposed an upper bound approach based on a translational failure mechanism. The
vertical slice techniques, which are often used in traditional limit equilibrium
approaches, are employed, and the force equilibrium is satisfied for all individual slices.
Two extreme kinematical solutions which neglect the interslice strength or fully utilize
Chapter 2: Literature Review

- 13 -
the interslice strength of the soil are then obtained. The traditional limit equilibrium
solutions of slices with a proper implicit assumption of failure mechanism can fall into
the range of these two extremes. Donald and Chen (1997) presented an upper bound
method on the basis of a multi-wedge failure mechanism, and the sliding body was
divided into a small number of discrete blocks. This approach was subsequently
employed by Wang (2001) and Wang et al. (2001) to examine the influence of a non-
associated flow rule on the safety factor. Some researchers have tried to use the finite
element method to obtain the upper bound solution for structures and geotechnical
problems (Anderheggen and Knopfel, 1972; Bottero et al., 1980; Sloan, 1988, 1989;
Sloan and Kleeman, 1995; Kim et al., 2002; Loukidis et al., 2003). A new upper bound
formulation using the rigid finite element method was presented by Chen (2004) and
Chen et al. (2003b, 2004, 2005a). The kinematically admissible velocity fields are
constructed by rigid elements, and this formulation render the limit analysis of slope
stability suitable to be conducted for different complex conditions, such as slopes with
complex geometries, soil profiles, groundwater conditions, and complicated loadings.

The application of the upper bound approach in 3D slope stability analysis was first
conducted by Giger and Krizek (1975, 1976), who analyzed the stability of a vertical cut
with a variable corner angle for cases with and without concentrated surcharge loading.
Michalowski (1989) presented an upper bound formulation for three-dimensional
analysis of locally loaded slopes. The through-toe and above-toe failure mechanisms are
considered, with energy dissipated along planar velocity discontinuities. Since this
method is limited to slopes with homogeneous soil, Farzaneh and Askari (2003)
improved and extended this method to non-homogeneous 3D slopes. Chen et al. (2001a,
2001b) proposed another 3D upper bound approach which is extended from the
corresponding 2D approaches by Donald and Chen (1997). In most of the 3D upper
Chapter 2: Literature Review

- 14 -
bound methods, the column techniques which are usually used in 3D LEM are
employed to construct the kinematically admissible velocity field. The vertical columns
are used by Michalowski (1989) and Farzaneh and Askari (2003), while the non-vertical
columns are used by Chen et al. (2001a, 2001b). The finite element method has also
been also used by some researchers to obtain 3D upper bound solution. Lyamin and
Sloan (2002b) proposed a new upper bound scheme using linear finite elements and
non-linear programming, and this approach is shown to be efficient for large scale three-
dimensional stability problems. Chen et al. (2005b) presented a three-dimensional upper
bound approach on the basis of the rigid finite element method, and a special sequential
quadratic programming algorithm is employed to obtain the optimal solution.

2.1.4 Strength reduction method
In recent decades, there are great developments of the SRM for slope stability analysis.
The general procedure of SRM analysis is the reduction of the strength parameters by
the factor of safety while the body forces due to weight of soil and other external loads
are applied until the system cannot maintain a stable condition. This procedure can
determine the safety factor within a single framework for both two- and three-
dimensional slopes. The main advantages of SRM are as follows: (1) the critical failure
surface is found automatically from the application of gravity loads and/or the reduction
of shear strength; (2) it requires no assumption on the interslice shear force distribution;
and (3) it is applicable to many complex conditions and can give information such as
stresses, movements, and pore water pressures. One of the main disadvantages of the
SRM is the long solution time required to develop the computer model and to perform
the analysis. With the development of computer hardware and software, 2D SRM can
now be performed within a reasonable time span suitable for routine analysis and
design, and 3D SRM can also be conducted within a tolerable time span. This technique
Chapter 2: Literature Review

- 15 -
is also adopted in several well-known commercial geotechnical finite element or finite
difference programs. In strength reduction analysis, the convergence criterion is the
most critical factor for the assessment of the factor of safety. Depending on the choice
of the program, different criteria for the ultimate state have been used in practice: (1)
maximum number of iteration is reached; (2) formation of a continuous failure
mechanism; (3) sudden change in the displacement for some selected points. For simple
problems, there are no major differences between these criteria, while major differences
may be obtained by different convergence criteria for some special cases.

The strength reduction method was early used in 2D slope stability analysis by
Zienkiewicz et al. (1975). Good agreement with slip circle solutions is obtained from an
idealized homogeneous embankment analysis. It is interesting to find that the associated
and non-associated assumptions have little influence on the results from a composite
embankment analysis.

Naylor (1982) applied the SRM to 2D problem and the failure was determined by the
development of very large displacements of protruding points on the failure mass.
Donald and Giam (1988) conducted SRM analysis in which the nodal displacement was
also used to assess the ultimate failure state. Donald and Giam (1988) states that it is
preferable to plot displacement curves for a number of nodes within the potential failure
region, and the nodes located in the toe region should be used if the failure zone is
unknown. The precision of the factor safety is found to be influenced by several factors,
such as the selection of constitutive model of soil, the location of the node selected for
plotting the displacement curve, tolerance for nonlinear analysis, the type of element
and the size of the discretized mesh.

Chapter 2: Literature Review

- 16 -
Brinkgreve and Bakker (1991) conducted SRM analysis for a river embankment in the
tidal zone and for a building trench supported by sheet-pile wall. An arc-length control
technique was used for the nonlinear equation solver to overcome the snap-through
problem associated with the ultimate limit state which will be followed by an
apparent negative stiffness matrix beyond the ultimate limit state. This technique
makes the analyzing robust as failure needs not be related with a non-converging
iterative procedure. Song (1997) also applied the arc-length control technique in the
SRM analysis, and good performance was obtained for two examples. A dam is
considered in the first example, and the factor of safety obtained by SRM agrees well
with the result obtained by the LEM. A road embankment reinforced with geotextiles is
studied in the second example, and it demonstrates the wide applicability of the SRM.
The arc-length control technique is now incorporated in the commercial finite element
software PLAXIS to obtain reliable collapse loads for load controlled calculations.

Matsui and San (1992) conducted SRM analysis in which the slope failure was defined
according to the shear strain failure criterion, and the hyperbolic nonlinear elastic soil
model was used for practical purpose. Ugai and Leshchinsky (1995) conducted 3D
SRM analysis for vertical cuts, which included a pseudo-static seismic force
component. Both the factors of safety and their corresponding slip surfaces determined
by the SRM demonstrate good agreement with the results from the LEM.

Griffiths and Lane (1999) conducted a detailed description of the SRM analysis and
carried out an extensive comparison against limit equilibrium solutions for several
typical slope examples. The failure is defined by the algorithm that convergence is not
achieved within a user-specified maximum number of iterations, and this implies that no
stable stress distribution, which satisfies both the Mohr-Coulomb failure criterion and
Chapter 2: Literature Review

- 17 -
the global equilibrium, can be found. The nodal displacement has a drastic increase at
the failure state. It is stated that the finite element method accompanied with an elastic-
perfectly plastic (Mohr-Coulomb) stress-strain relation can be a reliable and powerful
approach for calculating the factor of safety of slopes. It is also suggested that the SRM
should be a more powerful alternative to the traditional LEM.

Dawson et al. (1999) conducted an extensive simulation for a homogeneous
embankment with respect to a wide range of slope angles, soil friction angles and pore
pressure coefficients to assess the accuracy of the strength reduction technique. Factors
of safety were calculated using the finite difference software FLAC. The failure is also
defined by the non-convergence criterion, which is represented by the nodal unbalanced
force in FLAC. It is found that there is no ambiguity to determine the critical safety
factor, since the unbalanced force has a drastic change at the failure state, and this
phenomenon is due to the use of an elastic-perfectly plastic constitutive model, which
has a sharp transition from elastic to plastic behavior. Dawson et al. (1999) also points
out that if a model which shows smooth transition from elastic behaviour to plastic
behaviour is used (such as the hyperbolic constitutive model which was used by Matsui
and San 1992), determination of the failure state will be more difficult and will be more
sensitive to the tolerance in the nonlinear analysis. When the numerical mesh is fine
enough, the safety factors obtained by SRM are usually found to differ within a few
percent from the upper bound limit analysis results.

Zheng et al. (2005) also conducted an extensive SRM analysis and its application in the
slope, tunnel, and ultimate bearing capacity of foundations using program ANSYS.
Different soil models are considered for SRM analysis, and the results are compared
with limit equilibrium methods (Spencer approach) with good agreement.
Chapter 2: Literature Review

- 18 -
Griffiths and Marquez (2007) conducted strength reduction analysis for several 3D
slope examples. Both vertical and inclined boundaries were considered in the analysis to
investigate the constraint effect of slopes with finite length. Non-symmetric slopes with
weak soil surrounded by stronger soil were also analyzed. Besides, one example
analyzed by Baligh and Azzouz (1975) and two examples analyzed by Zhang (1988)
were also discussed.

Deng et al. (2007) have applied strength reduction method in assessing three-
dimensional stability of a pre-existing landslide with multiple sliding directions. Shukha
and Baker (2008) have employed the strength reduction approach to solve pseudo-static
stability problems in a frictional-cohesive material. Gurocak et al. (2008) have applied
strength reduction method in rock slope stability analysis using two-dimensional finite
element program Phase2 and the Hoek-Brown constitutive model.

2.2 Stability analysis of soil nailed slope
Soil nailing is a useful, economic technique for the construction of new steep cuts or the
strengthening of existing slopes. The essential concept of soil nailing is reinforcing the
slope with closely spaced inclusions to increase the stability. When the soil movement
is induced by excavation for cut slopes or by natural environment changes for existing
slopes, the resistant tension force is generated in the soil nail and is transferred into the
soil by the friction mobilized at the soil-nail interface.

Bruce and J ewell (1986) summarized the main characteristics and evolution history of
soil nailing technique, and revealed the potential benefits of soil nailing system. The
cost of soil nailing was found to be 10% to 30% less than that of the anchored
diaphragm or Berlin wall. The idea of the soil nailing technique stems from the mining
Chapter 2: Literature Review

- 19 -
engineering, in which the fully bonded steel inclusions combined with shotcrete are
used to provide immediate and effective support for rock tunneling system. It is
demonstrated that this technique can be employed in less competent materials. The
application of soil nailing in soils is successful in France, Germany and America in the
1970s, and this technique is employed for stabilizing excavations and slopes. Soil
nailing technique has been used in Hong Kong as a permanent slope stabilizing method
since the late 1980s.

In this section, the literature about soil nailing technique and its application is surveyed
and the review includes four parts. The first part discusses the three kinds of nail
resistance. The second part describes various failure modes of soil nailed slope. The
third part presents different design methods of soil nailed slope. The fourth part is
devoted to full scale and small scale tests of soil nailed slope.

2.2.1 Soil nail resistance
Soil nails have tensile resistance, shear resistance, and bending resistance. It is accepted
by many researchers that the axial tensile resistant force of the nail is the main
component in maintaining stability of a soil nailed slope, and the contribution from the
bending stiffness is very small unless the nails are installed nearly perpendicular to the
failure surface. If the nail is installed nearly horizontally, the effect of nail bending will
be very small. If the nail is installed at large angle below the horizontal direction and the
orientation is almost normal to the slip surface, more bending moment of the nail will
be mobilized, however, the bending stiffness still has only a modest improvement to the
slope stability compared with the contribution from the tensile resistance (J ewell and
Pedley, 1992). It is stated by Byrne et al. (1998) that the contribution to slope stability
by nail bending resistance is nearly one order of magnitude less than the contribution by
Chapter 2: Literature Review

- 20 -
the tension resistance. In addition, the bending resistance is not fully mobilized until
displacements are an order of magnitude larger than those required to generate
maximum tension resistance. Pedley and his colleagues (Pedley, 1990; Pedley et al.,
1990a, 1990b) maintain that the favorable effects from bending resistance will be the
post failure condition, therefore, the potential benefit of nail bending stiffness should be
neglected and not considered in design. There are however some different views on the
influence of the bending stiffness of the nail. Schlosser (1991) states that the effect of
the bending resistance can be either favorable or unfavorable which depends on the
behaviour of the soil nailing system, therefore, the bending and shear resistances have to
be taken into account in the design of soil nailing slopes. J uran et al. (1990, 1992) assert
that the nail bending stiffness has a considerable influence on the failure mechanism and
the mobilization of the bending stiffness may result in a remarkable decrease of the
slope stability.

Although there are some controversy on the influence of shear resistance and bending
resistance for soil nail, it is generally accepted that the principal resistance remains the
tensile force developed in the nails. Tensile force is generated in the soil nails primarily
as a result of the cohesive and frictional interaction between the nail and the soil.
Therefore, the nail-soil interface resistance (pullout resistance) becomes a chief
parameter which controls the design and stability evaluation of soil nailed slope. The
soil-nail pullout resistance is influenced by many factors, such as the normal stress
acting on the nail surface, the shear strength of the soil, the roughness of the nail
surface, the nail diameter, soil dilatancy and nail installing methods. Many researchers
have investigated the soil nail pullout resistance by analytical or empirical methods and
field or laboratory testing (Schlosser and Guilloux, 1981; Cartier and Gigan, 1983;
J ewell, 1990; Luo et al., 2000, 2002; Chu, 2003; Hong et al., 2003; J unaideen et al.,
Chapter 2: Literature Review

- 21 -
2004; Chu and Yin, 2005; Chai and Hayashi, 2005; Su, 2006; Pradhan et al., 2006; Su et
al., 2007; Li et al., 2008). In these investigations, one of the major concerns is whether
the pullout resistance is dependent on the depth or not (or overburden pressure).
Schlosser and Guilloux (1981) maintain that the pull-out resistance is independent of
the depth since the increase of the effective vertical stress offsets the decrease of the
apparent friction coefficient. J ewell (1990) suggests an equation in which the pullout
resistance is related to the depth. Su (2006) states that the pullout resistance is greatly
influenced by the dilatancy of the soil, and the peak pullout shear resistance is not
directly related to the overburden pressure.

2.2.2 Failure modes of soil nailed slope
There are three failure modes for reinforced slopes which are shown in Figure 2.1 (Elias
et al., 2001): internal failure where the slip surface passes through the reinforcing
inclusions; external failure where the slip surface passes behind and below the
strengthened soil mass; compound failure where the slip surface passes behind and
through the strengthened soil mass.

External failures include sliding failure, bearing failure, overturning failure and external
overall failure. The primary reasons for these failures are the insufficient length of nails
or the presence of weak soil at the slope base. In the stability analysis for slope with
external failure modes, the soil nailed slope mass is usually treated as a block. The
equilibrium of this block is established to compute the factor of safety by considering
the resistant forces exerting on the slip surface (Lazarte et al., 2003).

Internal failures, which are shown in Figure 2.2 (Byrne et al., 1998), include face
failure, pullout failure and tensile failure of nail tendon. The strength of the facing
Chapter 2: Literature Review

- 22 -
system in Figure 2.2(a) is not high, but the tensile strength of the nail is strong which
prevents the slope from nail tensile failure, and the nail length is very long which
prevents the slope from pullout failure. The facing failure is thus the most probable
collapse mode, and the active zone which is located between the facing and the failure
surface, slips off the front of the soil nail. The nail length in Figure 2.2(b) is limited, but
both the strength of the facing system and the nail are high, which prevent the slope
from facing failure and nail tensile failure. Pullout failure therefore occurs and the nail
slides out from the resistant zone, which is behind the failure surface. The tension
strength of the nail in Figure 2.2(c) is modest, but the facing system is strong and the
nail length is very long. The most likely failure mode for this case is hence the nail
tensile failure.

2.2.3 Design methods of soil nailed slope
The existing design approaches for soil nailed slope can be generally categorized into
two major types (ASCE 1997): limit equilibrium design methods which are essentially
based on the widely used solutions of modified slope stability analyses and are
employed to assess the overall stability of the soil nailing system by considering the
contribution of the nails; working stress design methods which are used to calculate the
mobilized nail forces at the service state and assess the local stability at each level of
nails.

Limit equilibrium methods have been developed by several investigators and these
methods include the Davis method (Shen et al., 1981a, 1981b), the German method
(Stocker et al., 1979; Stocker and Riedinger, 1990), and the French method (Schlosser,
1982, 1991). For the Davis method, only tensile resistance and pullout capacity of the
nail is considered and a parabolic sliding surface passing through the toe of the slope is
Chapter 2: Literature Review

- 23 -
assumed. The tensile forces mobilized in the nails are divided into tangential and
normal components along the potential slip surface and are added to the resistant forces
to determine the safety factor of the slope. For the German method, a bi-linear sliding
surface is usually assumed. Gassler (1988) extended this method, in which the slip
circle failure surface (rotation mechanism) is also considered, and it is found that this is
a relevant failure mode for practical design. Gassler (1988) also states that the rotation
of two rigid bodies may occur when very large surcharge loading is applied on the slope
crest. For the French method, two essential soil-nail interaction mechanisms are
considered, which include the soil-nail interface friction and the passive normal pressure
between the soil and the nail. This method takes into account the tensile resistance,
shearing resistance and bending stiffness of the nails. Four failure criteria are taken into
account: shear strength of the soil along the slip surface, strength of the nail, pullout
strength between soil and nail, lateral earth pressure on the nail.

In recent years, there is some development on the limit equilibrium method for soil nail
design. Sabahit et al. (1995) presented a design approach for soil nailed slopes in
combination with the modified J anbus generalized limit equilibrium method of slices.
The inclinations of nails and distribution of nail forces are regarded as main design
parameters and are used to obtain the minimum total nail force required to achieve a
desired value of safety factor. Sheahan and Ho (2003) proposed a trial wedge approach
to analyze the stability of soil nailing slopes. This method aims to provide a relatively
simple and inexpensive procedure for preliminary or complementary design of soil
nailed slope. Yuan et al. (2003) presented a limit equilibrium approach and it was
employed for reliability analysis of soil nailed slope. In this approach, the inter-slice
forces are calculated by recursion and the force equilibrium of the last boundary slice is
Chapter 2: Literature Review

- 24 -
satisfied by iteration. Patra and Basudhar (2005) developed a generalized method for
optimal design of soil nailing slopes.

Working stress design methods can be generally categorized into three major types
(ASCE, 1997): empirical design earth pressure diagrams, kinematical limit analysis, and
finite element/finite difference analysis.

Empirical design earth pressure diagrams are slightly modified from the earth pressure
design diagram proposed by Terzaghi and Peck (1967) for the design of braced
excavation, and are used to compute nail forces (J uran and Elias, 1987). This method
can only be employed for slopes with simple geometry and is not suitable to evaluate
the influence of various design parameters such as the inclination of the facing, rigidity
and inclination of the nails, surcharge loading, and heterogeneity of soil property.

Kinematical limit analysis for soil nailed slope design was proposed by J uran et al.
(1990). The upper bound limit analysis solution is employed in this design approach.
The constructed kinematically admissible failure mechanism is based on the perceived
failure mode in model tests. The local stability at the level of each nail can be assessed
and the nail forces can be calculated by this method. The major design parameters
(inclination and bending stiffness of nails, embankment slope, facing inclination, and
soil shear strength property) are taken into account in the formulation of this approach
to evaluate the slope stability. The influence of these parameters on the magnitude and
location of the maximum nail forces can also be assessed. In this method, the internal
failure criteria include the pullout failure, tensile failure and bending failure.

Chapter 2: Literature Review

- 25 -
The two-dimensional finite element or finite difference method for the analysis of nailed
slope has been used by Shen et al. (1981b), J uran et al. (1985), Plumelle et al. (1990),
Thompson and Miller (1990), Choukier (1996), Unterreiner et al. (1997), Murthy et al.
(2002), Babu et al. (2002), and Cheuk et al. (2005). Although two-dimensional analysis
provides valuable insight into the behaviours of nailed slope, the effect of soil-nail
interaction is not adequately considered. Some researchers adopt three-dimensional
finite element method to soil nailed slope analysis (Tabrizi et al., 1995; Smith and Su,
1997; Briaud and Lim, 1997; Zhang et al., 1999; Yang and Drumm, 2000). These
researchers focus mainly on the deformation, soil-nail interaction, and comparison of
the finite element predictions with observed behaviour of nailing slope, while the factor
of safety is seldom considered.

2.2.4 Full scale and small scale test of soil nailing slope
Some full scale and small scale model tests have been conducted to investigate the
behaviour of soil nailing system. Full scale test can provide the best representation of
the real behaviour of soil-nailed system, but it is very expensive and time consuming to
conduct these tests. There are thus only several full scale tests reported in the literature.
The first full scale test was the German research and development project
Bodenvernagelung which took place from 1975 until 1980 (Gassler and Gudehus,
1981; Gassler, 1992; Gassler, 1993). Seven fully instrumented field tests on nailed walls
were involved in this project, and the whole research scheme included the following
aspects: study the failure mechanisms of the nailed walls by theoretical analysis;
investigate the performance of nailed walls at limit equilibrium state by model tests;
develop drilling and construction technology in soils by implementation of seven full
scale nailed walls; perceive the behavior of nailed walls during the building process,
under service loading and at limit state by instrumentation in the tests.
Chapter 2: Literature Review

- 26 -
The second large full scale test was the French National Project Clouterre carried out
from 1986 to 1990 (Plumelle et al., 1990; Schlosser et al., 1992; Clouterre, 1993). Three
well instrumented experimental soil nailed walls were involved in this project. The
failure mode of the first testing wall was tensile failure of nail induced by partial
saturation of the soil from the top of the wall. The failure of the second testing nailed
wall was caused by increasing the height of the excavation. The collapse of the third
testing nailed wall arose from progressive shortening of the lengths of the nails.
Recently, a full scale test on a well instrumented soil-nailed loose fill slope was
conducted in Hong Kong (Li, 2003; Li et al., 2008). External concrete blocks are added
on the slope crest to induce the failure for wet and dry condition. It is found that soil
nailing combined with a surface grillage is a prospective approach to improve the
stability of loose fill slopes.

Though small scale or reduced scale test is not a prototype test, this experiment can still
provide good insight into the soil nailing behaviour and is much easier to be conducted
than full scale test. Stocker et al. (1979) reported some small scale tests, which were
carried out to investigate the bearing performance and failure mechanisms of nailed
retaining structures, and these tests were part of the German project
Bodenvernagelung. J uran et al. (1984) carried out some small scale model tests to
examine the performance of nailed soil and reinforced earth retaining structures, and to
study the nail tensile failure mechanism. Kitamura et al. (1988) conducted vertical
loading tests for nailed slope models to investigate the influence of the steel bar
reinforcement in a sandy slope. Hayashi et al. (1992) carried out a series of model tests
to evaluate the function of steel bar reinforcement in a cut-off slope with a thin deposit
layer. Kim et al. (1995) conducted small scale model tests to study the influence of
external loading on the failure mechanism of soil nailed slopes. Kim et al. (1996)
Chapter 2: Literature Review

- 27 -
performed model tests to examine the failure mechanism of soil nailed wall, and the
results show that bilinear or log-spiral lines can be good characterization for the slip
surface of soil nailed wall. Raju et al. (1997) conducted a series of model tests to
investigate the performance of a nailed wall controlled by its self weight.

2.3 Stability analysis of slopes reinforced with piles
Installation of piles to improve slopes stability has been demonstrated to be an effective
method (DAppolonia et al., 1967; De Beer and Wallays, 1970; Ito and Matsui, 1975;
Fukuoka, 1977; Wang et al., 1979; Ito et al., 1981, 1982; Reese et al., 1992). A number
of approaches have been employed to evaluate the performance and design of the piles
which are used as reinforcement in slopes, and these methods are summarized as
follows.

Ito and Matsui (1975) proposed a theoretical formulation to analyze the growth
mechanism of lateral force exerting on stabilizing piles in a row, and the influence of
pile spacing was considered. An accurate estimation about the lateral force is very
important in the design of the reinforcing piles, and the magnitude of the force is related
to the movement of the soil surround the pile. The formulation presented by Ito and
Matsui (1975) estimates a value of the lateral force in the initial state of landslide
movement, and this value is less than the force in the ultimate limit state of landslide
movement. Two kinds of formulation were derived on the basis of the theories of plastic
deformation and plastic flow. The expression by the latter theory assumes a visco-
plastic flow occurring around the piles, and this formulation is suitable for very soft
soil. The expression by the former theory is applicable for common soils, and some of
the assumptions used in this formulation by Ito and Matsui (1975) are: when the soil
layer deforms, two sliding surfaces occur on the two sides of the pile which have an
Chapter 2: Literature Review

- 28 -
angle of (/4+/2) with the axis normal to the pile row plane; the soil is in a state of
plastic equilibrium only in the soil body around the pile where the Mohr-Coulomb yield
criterion applies; the soil layer is in a plane-strain condition in the direction of depth; the
frictional forces on the sliding surfaces of the two sides of the pile are neglected when
the stress distribution in the soil body around pile is considered; the piles are rigid. Ito et
al. (1981, 1982) subsequently developed a design methodology for a pile reinforced
slope based on the above approach. Hassiotis et al. (1997) also presented an approach
for the design of slopes strengthened by a single row of piles in which the plastic state
theory developed by Ito and Matsui (1975) was also used to estimate the pressure acting
on the piles.

The boundary element method (Poulos, 1973; Poulos and Davis, 1980; Lee et al., 1991;
Chen and Poulos, 1997) has also been employed to assess the passive pile response
considering the soil moving through the pile. On the basis of this method, Lee et al.
(1995) and Poulos (1995) have developed an approach for the design of piles to
reinforce slopes. The design procedure by Poulos (1995) is as follows: (1) calculating
the total shear force required to improve the slope stability to a desired standard; (2)
assessing the maximum shear force that each pile can offer to prevent potentially
unstable soil mass from sliding by pile-soil interaction analysis using boundary element
method; and (3) determining the type and number of piles, and the optimal pile position.

Chow (1996) presented a numerical method for the analysis of pile stabilized slope. In
this method, the piles are simulated by beam finite elements and the soil is simulated by
a hybrid approach, in which the subgrade reaction modulus is used to model the soil
response, and the elastic theory is used to model the interaction between the pile and the
soil.
Chapter 2: Literature Review

- 29 -
Yamagami et al. (2000) presented a limit equilibrium design method for slopes
reinforced with a row of piles. In this method, two individual failure surfaces are
assumed on the upper part and lower part of the pile. Based on the presumed slip
surface, the forces exerting on the stabilizing piles can be evaluated according to the
prescribed safety factor.

The kinematic approach of limit analysis was employed by Ausilio et al. (2001) to
conduct stability analyzing for soil slopes strengthened with piles. The formulation was
derived based on the force required for improving the slope stability to a prescribed
standard, and the optimal pile position was assessed. It is shown that the optimum pile
location is usually near to the slope toe, and the critical slip surface for slope reinforced
with pile is deeper than an unreinforced slope.

A new subgrade reaction formulation was developed by Cai and Ugai (2003) to evaluate
the response of flexible piles in landslides, and the dimensionless design charts were
presented on the basis of this new formulation. The influence of the linear movement of
the sliding layer was taken into account in this formulation.

The finite element and finite difference method have also been employed in piled slope
analysis by some researchers. Cai and Ugai (2000) have conducted stability analysis for
slope reinforced with piles by three-dimensional finite element method using shear
strength reduction technique, and the influence of the pile spacing, pile head conditions,
bending stiffness, and pile locations on the slope stability were considered. Won et al.
(2005) have analyzed the same slope model as Cai and Ugai (2000) by three-
dimensional finite difference code FLAC3D, and the shear strength reduction technique
was also used to obtain the factor of safety. Both the finite element results by Cai and
Chapter 2: Literature Review

- 30 -
Ugai (2000) and finite difference results by Won et al. (2005) were compared with those
obtained by limit equilibrium method, where the reaction force of the piles was
determined by equation from Ito and Matsui (1975). The factors of safety by the SRM
are slightly larger than those by the limit equilibrium method, but the slip surface by the
SRM is much deeper which is determined by the maximum point of shear force in the
pile. The finite deference method has also been used by Ng and Zhang (2001), Ng et al.
(2001) and Zhang et al. (2004) for laterally loaded sleeved piles on slope stability. A
series of 3D analyses were conducted to study the behavior of laterally loaded piles and
the effects of unsleeved and sleeved piles on the stability of a cut slope, and the SRM
was used in evaluating the slope stability. Chen and Martin (2002) have used finite
difference program FLAC to investigate the arching effect and mobilization mechanism
of the resistant force in passive pile groups when lateral soil movement is induced.
Martin and Chen (2005) have also used this program to assess the behavior of piles
installed in an embankment slope with a translational failure mode. Pan et al. (2002)
and Miao et al. (2006) have used finite element program ABAQUS to investigate the
passive pile behavior.

2.4 Stability analysis of slopes with water flow
Water-induced slope failure is a common geotechnical problem. The stability of the
slope is reduced because of the seepage force. For slope stability analysis with water
flow, the widely used method is still the limit equilibrium method. The inclusion of
seepage forces in the limit equilibrium method has some confusion in the past. In the
traditional limit equilibrium method, the boundary water forces with total weights are
usually used and the water pressures enter into the base normal/shear force calculation
but not along the sides between slices. Turnbull and Hvorslev (1967) viewed that the
traditional method yields unreasonable results for high value of pore-pressure and
Chapter 2: Literature Review

- 31 -
suggested only the effective stress should be resolved in a direction normal to the slip
surface. Greenwood (1983, 1985) and King (1989) introduced some effective-stress
methods of slices which including interslice water forces. Actually all these methods do
not have great difference since we can use either boundary water forces with total
weights or seepage forces with submerged weights and these two methods should give
the same answer theoretically. The uses of the boundary forces and total weights are
usually more convenient and are widely accepted and employed in practice.

For slope stability analysis with water flow, the pore water pressure should be properly
determined. In limit equilibrium method, the piezometric line is most commonly used
for determining the pore-water pressure distribution. The vertical length from the slice
base mid-point up to the piezometric line is calculated, and this length times the unit
weight of water will give the pore-water pressure at the slice base. In a slope with water
flowing from upper part to lower part, the conventional assumption about the
piezometric surface is usually not accurate. In some commercial software, such as
SLOPE/W (2004), a correction factor is introduced to more correctly determine the
pore-water pressure. Another method, which is also employed in some commercial
software is the integration of the limit equilibrium method with the finite element
program and to adopt the pore pressure from the flow-net. This integrated procedure has
been employed in slope stability analysis with seepage flow by Ng and Shi (1998),
Gasmo et al. (2000), Tsaparas et al. (2002), Kim et al. (2004), and Rahardjo et al.
(2007).

Limit analysis method was used by Kim et al. (1999) for the analysis of slope subjected
to pore water pressures. Both the statically admissible stress fields for the lower bound
solution and kinematically admissible failure mechanisms for the upper bound solution
Chapter 2: Literature Review

- 32 -
were constructed by three-noded linear triangular finite elements. The influence of the
water pressure were taken into account and combined into the finite element analysis.

Another approach, which is called finite element stress method, was proposed by
Fredlund et al. (1999). In this method, a finite element stress analysis is combined with
a limit equilibrium analysis. The stress distribution in the slope can first be computed by
finite element analysis. The stress distribution from the finite element analysis is then
introduced into the LEM analysis for calculating the normal stress and mobilized shear
stress along any given failure surface. The factor of safety can be determined as the ratio
of the summation of the available resistant shear force to the summation of the
mobilized shear force along the selected slip surface. If this approach is applied for
slope analysis with water flow, couple finite element analysis should be conducted to
obtain the effective stress distribution. Cho and Lee (2001) employed this method to
calculate the safety factor for an unsaturated slope suffering from rainfall infiltration.

Strength reduction method can also be used in slope stability analysis with water flow.
The pore-pressure can be computed by the seepage or consolidation analysis, the
effective stress is then used to conduct shear strength reduction analysis. This process
can easily be completed in a single finite element or finite difference software, and it
does not require coupling of finite element analysis and limit equilibrium analysis. This
approach has been employed by Cai et al. (1999), Cai and Ugai (2004) on slope stability
analysis under rainfall.




Chapter 2: Literature Review

- 33 -


Figure 2.1 Failure modes for reinforced soil slopes (after Elias et al., 2001)


Figure 2.2 Potential soil nail wall internal failure modes (after Byrne et al., 1998)
Chapter 3: Two-dimensional Slope Stability Analysis by Limit Equilibrium and Strength Reduction Methods

- 34 -

CHAPTER 3: TWO-DIMENSIONAL SLOPE STABILITY
ANALYSIS BY LIMIT EQUILIBRIUM AND STRENGTH
REDUCTION METHODS

3.1 Introduction
In this chapter, the factors of safety and the locations of critical failure surfaces obtained
by the limit equilibrium method and strength reduction method are compared for
various 2D slopes. The study in this chapter will provide a good investigation on the
two most important slope stability methods, and the advantages and limitations of these
two methods are investigated under different conditions. For SRM analysis, the FOS of
a slope is defined as the reduction factor by which the original soil shear strength
parameters should be reduced to make the slope to reach the critical failure state. This
definition is the same as the one employed in the limit equilibrium method. Thus the
reduced shear strength parameters
f
c and
f
are given as follows:
FOS / c c
f
= (3.1)
) FOS / arctan(tan =
f
(3.2)
There is another definition of the FOS for SRM in which the load or gravity is increased
by a certain factor to bring the slope to the critical failure state, and this definition is
different from that in the traditional LEM. In the present study, the first definition is
used in all the SRM analysis, and since this definition is the same as the one used in
LEM, the comparison of the FOS between SRM and LEM will be reasonable.

Chapter 3: Two-dimensional Slope Stability Analysis by Limit Equilibrium and Strength Reduction Methods

- 35 -
Many researchers have compared the results between the SRM and LEM and have
found that generally the two methods will give similar FOS. Most of the studies are,
however, limited to homogenous soil slopes and the geometry of the problems is
relatively regular with no special features (e.g. the presence of a thin layer of soft
material or special geometry). Furthermore, there are only limited studies which
compare the critical failure surfaces from the LEM and SRM as the FOS appears to be
the primary quantity of interest. In this chapter, the two methods are compared under
different conditions and both the FOS and the locations of the critical failure surfaces
are considered in the comparisons. In the present study, both a non-associated flow rule
(SRM1 and soil dilation angle=0) and an associated flow rule (SRM2 and soil dilation
angle=friction angle) are applied for Mohr-Coulomb model in the SRM analyses. To
define the critical failure surface from the SRM, both the maximum shear strain rate
(the rate of variation of the shear strain at the failure state) and the maximum shear
strain increment (the accumulated total shear strain at the failure state) definition can be
used. It has been found that these two definitions will give similar locations of the
critical failure surface under most cases for different computer programs. The failure
surface may also be defined by displacement distribution, but this criterion is usually
not very clear as compared with the shear strain criterion.

In the present study, the limit equilibrium method is considered using the Morgenstern-
Price method with f(x)=1.0 (equivalent to Spencer method). Krahn (2003) and
Abramson et al. (2002) have pointed out that f(x) may be critical for some special cases,
but this is generally not the case unless the problem is highly complicated. Besides
Spencers method, the Generalized Limit Equilibrium (GLE) method (Abramson et al.
2002) has also been tried and the results are also similar to those by the Morgenstern-
Price method. The differences of the FOS and the critical failure surfaces from f(x)=1.0
Chapter 3: Two-dimensional Slope Stability Analysis by Limit Equilibrium and Strength Reduction Methods

- 36 -
and f(x)=sin(x) are also found to be small for both the Morgenstern-Price method and
the GLE method in the present study. Hence, the discussions in this chapter are not
sensitive to the specific form of LEM. The software Slope2000 by Cheng is used for the
LEM analysis.

In performing the SRM analysis, many soil parameters and boundary conditions are
required to be defined which are absent in the corresponding LEM analysis. The
importance of the various parameters and the applicability of the SRM under several
special cases are considered in the following sections. For SRM analysis, except for
some special cases which are analyzed by other software, all the examples shown in this
chapter are analyzed by commercial software FLAC3D.

3.2 Stability analysis for a simple and homogeneous soil slope
Firstly, a homogeneous soil slope with a slope height equal to 6m and slope angle equal
to 45 (Figure 3.1) is considered. For the three cases in which the friction angle is 0,
since the critical slip surface is a deep-seated surface with a large horizontal extent, the
models are larger than the one as shown in Figure 3.1 and have a width of 40m and a
height of 16m. In the parametric study, different shear strength properties are used and
the LEM, SRM1 and SRM2 analyses are carried out. The cohesive strength c of the soil
varies from 2 kPa, 5 kPa and 10-20 kPa while the friction angle varies from 5, 15,
25 and 35 - 45. The density, elastic modulus and Poisson ratio () of the soil are kept
at 20 kN/m
3
, 14MPa and 0.3 respectively in all the analysis. As shown in Figure 3.1, the
size of the domain for the SRM analyses is 20m in width and 10m in height and there
are 3520 zones and 7302 grid points in the mesh for analysis. Based on limited mesh
refinement studies, it was found that the discretization as shown in Figure 3.1 is
Chapter 3: Two-dimensional Slope Stability Analysis by Limit Equilibrium and Strength Reduction Methods

- 37 -
sufficiently good so that the results of analyses are practically insensitive to a further
reduction in the element size. For the LEM, the Morgenstern-Price method, which
satisfies both moment and force equilibrium, is adopted and the critical failure surface is
evaluated by the modified simulated annealing technique as proposed by Cheng (2003).
The tolerance for locating the critical failure surface by the simulated annealing method
is 0.0001 which is sufficiently accurate for the present study.

From Table 3.1 and Figures 3.2 to 3.5, it is found that the FOS and critical failure
surfaces determined by the SRM and LEM are very similar under different
combinations of soil parameters for most cases except when =0. When the friction
angle is greater than 0, most of the FOS obtained by the SRM differ by less than 7.4 %
with respect to the LEM results except for case 16 (c=20kPa, =5) where the
difference is up to 13.2 %. When the friction angle is very small or zero, there are
relatively major differences between the SRM and LEM for both the FOS and the
critical slip surface (Table 3.1 and Figure 3.5). Based on Table 3.1 and Figures 3.2 to
3.5, some conclusions can be made as follows:

(1) Most of the FOS obtained from the SRM are slightly larger than those obtained from
the LEM with only few exceptions.

(2) The FOS from an associated flow rule (SRM2) are slightly greater than those from a
non-associated flow (SRM1), and this difference increases with increasing friction
angle. These results are reasonable and are expected. The differences between the two
set of results are, however, small because the problem has a low level of kinematic
constraint.

Chapter 3: Two-dimensional Slope Stability Analysis by Limit Equilibrium and Strength Reduction Methods

- 38 -
(3) When the cohesive strength of the soil is small, the differences in FOS between the
LEM and SRM (SRM1 and SRM2) are greatest for higher friction angles. When the
cohesion of the soil is large, the differences in FOS are greatest for lower friction
angles. This result is somewhat different from that of Dawson et al. (2000), who
concluded that the differences are greatest for higher friction angles when the results
between SRM and limit analysis are compared.

(4) The failure surfaces from the LEM, SRM1 and SRM2 are similar in most cases. In
particular, the critical failure surfaces obtained by SRM2 appear to be closer to those
from LEM than those obtained from SRM1. The critical failure surfaces from SRM1,
SRM2 and LEM are practically the same when the cohesive strength is small (it is
difficult to differentiate clearly in Figures 3.2, 3.3a, 3.3b, 3.4a, 3.4b), but noticeable
differences in the critical failure surfaces are found when the cohesive strength is high
(Figures 3.3d, 3.4d, 3.5a, 3.5b).

(5) The right end of the failure surface moves closer to the crest of the slope as the
friction angle of the soil is increased (which is a well known result). This behaviour is
more obvious for those failure surfaces obtained from SRM1. For example, for the five
cases where the cohesion of the soil is 2kPa (Figure 3.2), when the friction angle is 5,
15 and 25 degrees, the right end-point of the failure surface derived from SRM1 is
located to the right of the right end-point of the critical failure surface obtained from the
LEM. When the friction angle is 35, the right end-point of the failure surface obtained
by the SRM1 and LEM are nearly at the same location. When the friction angle is 45,
the distance of the right end-point derived from SRM1 is located to the left of the right
end-point derived from LEM.

Chapter 3: Two-dimensional Slope Stability Analysis by Limit Equilibrium and Strength Reduction Methods

- 39 -
(6) For SRM analyses, when the friction angle of soil is small, the differences between
the slip surfaces for SRM1 and SRM2 are greatest for smaller cohesion (Figure 3.3).
When the friction angle is large, the differences between the slip surface for SRM1 and
SRM2 are greatest for higher cohesion (Figure 3.4).

(7) It can also be deduced from Figures 3.2 to 3.5 that the potential failure volume of the
slope becomes smaller with increasing friction angle but increases with increasing
cohesion. This is also well known behaviour, as when the cohesive strength is high the
critical failure surface will be deeper.

Although there are some minor differences in the results between the SRM and LEM in
this example, the results from these two methods are generally in good agreement which
suggests that the use of either the LEM or SRM is satisfactory in general. However, an
interesting case is constructed in the next section where a limitation of the SRM is
demonstrated.

3.3 Stability analysis of a slope with a soft band
A special problem with a soft band has been constructed as it appears that similar
problems have seldom been considered previously. The geometry of the slope is shown
in Figure 3.6 and the soil properties are shown in Table 3.2 (In practice the elastic
modulus should not be the same for different soil. The elastic modulus used in this
section for different soil layer is assumed to be 14MPa for simplicity; and it will be
shown in the later sections that the influence of the elastic modulus on the SRM analysis
is very small, so the use of the elastic parameters should not have great effect on the
result). It is noted that c is zero and is small for soil layer 2 which has a thickness of
just 0.5m. The critical failure surface is obviously controlled by this soft band, and slope
Chapter 3: Two-dimensional Slope Stability Analysis by Limit Equilibrium and Strength Reduction Methods

- 40 -
failures in similar conditions have actually occurred in Hong Kong (for example, the Fei
Tsui Road slope failure in Hong Kong).

In order to consider the size effect (boundary effect) in the SRM, three different
numerical models are developed to perform the SRM using Mohr-Coulomb analysis and
the widths of the domains are 28m, 20m and 12m, respectively (Figure 3.7). In these
three SRM models, various mesh sizes were tried until the results were insensitive to the
number of elements used for the analysis. For example, when the domain size is 28m,
the FOS was found to be 1.37 (Table 3.3a) with 12000 elements, 1.61 with 6000
elements and 1.77 with 3000 elements using SRM1 analysis and the program Phase2.

Since the FOS for this special problem have great differences from those found using
the LEM, several well known commercial programs (Flac3D, Flac2D, Phase2, Plaxis)
are tried and very surprising results are obtained. For these four commercial programs,
Flac3D and Flac2D are finite difference codes, while Phase2 and Plaxis are finite
element codes, but the definition of SRM in these codes is practically the same. The
ultimate limit state is determined by non-convergence criterion in all these codes, and an
arc-length control technique is used in Plaxis to determine the ultimate limit state. The
locations of the critical failure surfaces from SRM for solution domain width of 12m,
20m and 28m are virtually the same. The local failures from the SRM, as shown in
Figure 3.8b, range from x=5m to x=8m and the failure surfaces are virtually the same
for the three different solution domains. The majorities of the critical failure surfaces lie
within layer 2, which has low shear strength, and are far from the right boundary. It is
surprising to find that different programs produce drastically different results (Table
3.3a) for the FOS even though the locations of the critical failure surface from these
programs are very similar. For the cases as shown in Figure 3.1, and other cases in a
Chapter 3: Two-dimensional Slope Stability Analysis by Limit Equilibrium and Strength Reduction Methods

- 41 -
latter part of this chapter, the results are practically insensitive to the domain size, while
the cases shown in Figure 3.6 are very sensitive to the size of domain for the programs
Flac3D (SRM1 and SRM2) and Phase2 (SRM2). Results from the Plaxis program
appear to not be sensitive to the domain size but are quite sensitive to the dilation angle
(which is different from the previous example). The SRM1 results from program Phase2
are also not sensitive to the domain size for SRM1, but results from SRM2 behave
differently. The FOS from Flac3D appear to be overestimated when the soil parameters
for the soft band are low, but the results from this program are not sensitive to the
dilation angle which is similar to all the other examples in the present study. For SRM1,
the results from Phase2 and Plaxis appear to be more reasonable as the results are not
sensitive to the domain sizes, while for SRM2, results from Plaxis may be better. It is
also surprising to find that Flac2D cannot give any result for this problem, even after
many different trials, but the program worked properly for all the other examples in this
study.

There is another interesting and important issue when SRM is adopted for the present
problems. For the problem with a 12m domain, Phase2 cannot provide a result with the
default settings and these settings (including the tolerance and number of iterations
allowed) are varied until convergence is achieved. The results of analysis for a 12m
domain with Phase2 are shown in Tables 3.4 and 3.5. It is observed that the number of
elements used for the analysis has a very significant effect on the factor of safety, which
is not observed for the cases in Table 3.1. The tolerance used in the nonlinear equation
solution also has a major impact on the results for this case. This is less obvious for
other cases considered in the present study.

Chapter 3: Two-dimensional Slope Stability Analysis by Limit Equilibrium and Strength Reduction Methods

- 42 -
Besides the special results shown above, the FOS from the 28m domain analysis
appears to be large for Flac3D and Phase2 when the strength parameters for the soil
layer 2 are low. In fact, it is not easy to define an appropriate factor of safety from SRM
analysis for this problem. If the cohesive strength of the top soil is reduced to zero, the
factor of safety can be estimated as 0.57 from the relation tan/tan, where is the
slope angle. It can be seen that for the LEM, the cohesive strength 20 kPa for soil 1
helps to bring the factor of safety to 0.927 and a high factor of safety for this problem is
not reasonable. Without the results from the LEM for comparison, it may be
unconservative to adopt the values of 1.64 (1.61) from the SRM based on Flac3D.

When the soil properties of the soft band are changed to c=10 kPa and =0, the results
of analyses are as shown in Table 3.3b. It is found that the critical failure will extend to
a much greater distance so that a 28m wide domain is necessary. The FOS from the
different programs are virtually the same, which is drastically different from the results
in Table 3.3a (the same meshes were used for Table 3.3a and Table 3.3b).

If the soil properties of soils 2 and 3 are interchanged so that the third layer of soil is the
weak soil, the FOS from SRM2 are 1.33 by FLAC3D for all three different domain
sizes. The corresponding factor of safety from the LEM is 1.29 from the Morgenstern-
Price analysis. The locations of the critical failure surface from the SRM and LEM for
this case are also very close, except for the initial portion, as shown in Figure 3.9a and
3.9b. It appears that the presence of a soft band with frictional material, instead of major
differences in the soil parameters is the actual cause for the difficulties in the SRM
analysis. Great care is required in the implementation of a robust nonlinear equation
solver for the SRM.

Chapter 3: Two-dimensional Slope Stability Analysis by Limit Equilibrium and Strength Reduction Methods

- 43 -
The problems as shown in Table 3.3a may reflect the limitations of commercial
programs rather than the limitations of the SRM, but they illustrate that it is not easy to
compute a reliable FOS for this type of problem using the SRM. The results are highly
sensitive to different nonlinear solution algorithms which are not clearly explained in
the commercial programs. Great care, effort and time are required to achieve a
reasonable result from SRM for this special problem and comparisons with the LEM are
necessary. It is not easy to define a proper factor of safety from the SRM alone for the
present problem as the results are highly sensitive to the size of domain and the flow
rule. In this respect, the LEM appears to be a better approach for this type of problem.

3.4 Local minimum in LEM
For the LEM, it is well known that many local minima may exist besides the global
minimum. This makes it difficult to locate the critical failure surface by classical
optimization methods. Comparisons of the LEM and SRM with respect to local minima
have seldom been considered in the past, but this is actually a very important issue
which is illustrated by the following examples. In the SRM, there is no local minimum
as the formation of the shear band will attract strain localization in the solution process.
To investigate this issue, an 11m height slope as shown in Figure 3.10 is considered.
The slope angle for the lower part of the slope is 45 while the slope angle for the upper
part of the slope is 26.7. The cohesion and friction angle of the soil are 10 kPa and 30,
respectively, and the density of the soil is 20 kN/m
3
.

The failure mechanism by the SRM is shown in Figure 3.11 and the FOS is 1.47 for
both non-associated flow and associated flow. The right end-point of the failure surface
is located to the right of the crest of the slope. The results derived from the LEM are
presented in Figure 3.12 (number of slices is 50). The global minimum factor of safety
Chapter 3: Two-dimensional Slope Stability Analysis by Limit Equilibrium and Strength Reduction Methods

- 44 -
is 1.383 but a local minimum FOS of 1.3848 is also found. The location of the failure
surface for the local minimum 1.3848 is very close to that from the SRM, and the
failure surface for the global minimum from the LEM is not the critical failure surface
from the SRM. Since the FOS for the two critical failure surfaces from the LEM are so
close, both failure surfaces are probable failure surfaces and should be considered in
slope stabilization. For the SRM, there is only one unique failure surface from the
analysis and another possible failure mechanism cannot be easily determined. Thus, the
SRM analysis may yield a local failure surface of less importance while a more severe
global failure surface remains undetected, as illustrated in the next example. This is
clearly a major drawback of the SRM as compared with the LEM.

Another interesting case which is worth discussion is constructed. Figure 3.13 shows a
relatively simple slope with a total height of 55m in a uniform soil. The soil parameters
are c=5 kPa and =30 while the unit weight is 20 kN/m
3
. The global minimum and
local minima are determined in accordance with the procedures of Cheng (2003) and
different boundaries for the left and right exit ends are specified in the study. Using the
LEM, the global minimum FOS is obtained as 1.33 (Figure 3.13a) but several local
minima are found with factors of safety in the range 1.38 to 1.42 as shown in Figures
3.13b-3.13e. From the SRM, only the factor of safety 1.327 is found which is similar to
the global minimum shown in Figure 3.13a. If slope stabilization is only carried out for
this failure surface, the possible failure surfaces shown by Figures 3.13d and 3.13e will
not be considered. Baker and Leshchinsky (2001) have proposed the concept of the
safety map, which enables the global minimum and local minima from the LEM to be
visualized easily, but it appears that construction of such a map using the SRM is
impossible. In this respect, the LEM is a better tool for slope stability analysis. It is
possible that the use of the SRM may miss the location of the next critical failure
Chapter 3: Two-dimensional Slope Stability Analysis by Limit Equilibrium and Strength Reduction Methods

- 45 -
surface (with a very small difference in the FOS but a major difference in the location of
the critical failure surface) so that the slope stabilization measures may not be adequate.
This interesting case has illustrated a major limitation of the SRM for the design of
slope stabilization works.

3.5 Influence of elastic modulus for SRM analysis
In order to investigate the influence of the soil elastic modulus in slope stability analysis
when the SRM is used, the four different models shown in Figures 3.14-3.17 are
considered. In all these models, the strength parameters are the same for every soil so
that there is actually only one soil in the LEM analysis. For the SRM analyses, different
elastic moduli are used for different soils and the results of the analyses, shown in
Tables 3.6-3.9, indicate that they are relatively insensitive to the elastic modulus of soil.
When the elastic moduli of soils 1 and 2 are interchanged, there is only a minor change
in the factor of safety. In fact, the effect of flow rule is slightly more important than the
elastic moduli in the present example. In general, the differences in both the FOS and
the locations of the critical failure surfaces are small between the LEM and SRM. For
case 2 in Table 3.9, the difference in the FOS between the LEM and SRM is significant
but the locations of the critical failure surfaces from the two methods (Figure 3.18) are
close to each other.

3.6 Discussion and Conclusion
In the present study, a number of interesting features of the SRM were highlighted
which are important for a proper analysis of a slope. While most research has
concentrated on the FOS between the LEM and SRM, the present works have compared
the locations of the critical failure surfaces from these two methods. In a simple and
homogenous soil slope, the differences in the FOS and locations of the critical failure
Chapter 3: Two-dimensional Slope Stability Analysis by Limit Equilibrium and Strength Reduction Methods

- 46 -
surfaces from the SRM and LEM are small and both methods are satisfactory for
engineering use. It is found that when the cohesion of the soil is small, the difference in
FOS from the two methods is greatest for higher friction angles. When the cohesion of
the soil is large, the difference in FOS is greatest for lower friction angles. With regard
to the flow rule, the FOS and locations of the critical failure surface are not greatly
affected by the choice of the dilation angle (which is important for the adoption of the
SRM in slope stability analysis). When an associated flow rule is assumed, the critical
slip surfaces from SRM2 appear to be closer to those from the LEM than those from
SRM1.

For the SRM, the effects of the dilation angle, the tolerance for nonlinear equation
analysis, the soil moduli and the domain size (boundary effects) are usually small but
still noticeable. In most cases, these factors cause differences of just a few percent and
are not critical for engineering use of the SRM. Since the use of different LEM methods
will also give differences in the factor of safety of several percent, the LEM and SRM
can be viewed as similar in performance for normal cases.

Drastically different results are determined from different computer programs for the
problem with a soft band. For this special case, the factor of safety is very sensitive to
the size of the elements, the tolerance of the analysis and the number of iteration
allowed. It is strongly suggested that the LEM be used to check the results from the
SRM. This is because the SRM is highly sensitive to the nonlinear solution algorithms
and flow rule for this special type of problem. Even though results from the program
Plaxis (and some results from the program Phase2) are insensitive to the domain size,
they are very sensitive to the flow rule which is a result not found in all the other
Chapter 3: Two-dimensional Slope Stability Analysis by Limit Equilibrium and Strength Reduction Methods

- 47 -
examples. The SRM has to be used with great care for problems with a soft band of this
nature.

The two examples with local minima for the LEM illustrate another limitation of the
SRM in engineering use. With the SRM, there is strain localization during the solution
and the formation of local minima is unlikely. In the LEM, the presence of local minima
is a common phenomenon, and this is a major difference between the two methods.
Thus, it is suggested that the LEM should be preformed in conjunction with the SRM as
a routine check.

Through the present study, two limitations of the SRM have been established: (1) it is
sensitive to nonlinear solution algorithms/flow rule for some special cases and (2) it is
unable to determine other failure surfaces which may be only slightly less critical than
the SRM solution but still require treatment for good engineering practice. If the SRM is
used for routine analysis and design of slope stabilization measures, these two major
limitations have to be overcome and it is suggested that the LEM should be carried out
as a cross reference. If there are great differences between the results from the SRM and
LEM, great care and engineering judgment should be exercised in assessing a proper
solution. There is one practical problem in applying the SRM to a slope with a soft
band. When the soft band is very thin, the number of elements required to achieve a
good solution is extremely large so that very significant computer memory and time are
required. Cheng (2003) has tried a slope with a 1mm soft band and has effectively
obtained the global minimum factor of safety by the simulated annealing method. If the
SRM is used for a problem with a 1mm thick soft band, it is extremely difficult to
define a mesh with a good aspect ratio unless the number of elements is huge. For the
SRM with a 500 mm thick soft band, about one hour of CPU time for a small problem
Chapter 3: Two-dimensional Slope Stability Analysis by Limit Equilibrium and Strength Reduction Methods

- 48 -
(several thousand elements) and several hours for a large problem (over ten thousand
elements) were required for the Phase2 program, while the program Flac3D required 1-
3 days (for small to large meshes). If a problem with a 1mm thick soft band is to be
modeled with the SRM, the computer time and memory required will be huge and the
method is not applicable for this special case. The LEM is perhaps better than the SRM
for these cases.

For the SRM, there are further limitations which are worth observing. Shukha and
Baker (2003) have found that there are minor but noticeable differences in the factors of
safety from Flac using square elements and distorted elements. The use of distorted
elements are however unavoidable in many cases. Furthermore, when both the soil
parameters c and are very small, it is well known that there are numerical problems
with the SRM. The failure surface in this case will be deep and wide and a large domain
is required for analysis. It has been found that the solution time is extremely long and a
well-defined critical failure surface is not well established from the SRM. For the LEM,
there is no major difficulty in estimating a factor of safety and the critical failure surface
under these circumstances.

The advantage of the SRM is the automatic location of the critical failure surface
without the need for a trial and error search. With the use of modern global optimization
techniques, the location of critical failure surfaces by a simulated annealing method, a
genetic algorithm or other methods is now possible and trial and error search with the
LEM is no longer required. While the LEM suffers from the limitation of an interslice
shear force assumption, the SRM suffers from being sensitive to the nonlinear solution
algorithm/flow rule for some special cases.

Chapter 3: Two-dimensional Slope Stability Analysis by Limit Equilibrium and Strength Reduction Methods

- 49 -
Griffith and Lane (1999) have suggested that a non-associated flow rule should be
adopted for slope stability analysis. As the effect of flow rule on the SRM is not
negligible in some cases, such as those involving a soft band, the flow rule is indeed an
issue for a proper slope stability analysis. It can be concluded that both the LEM and
SRM have their own merits and limitations, and the use of the SRM is not really
superior to the use of the LEM in routine analysis and design. Both methods should be
viewed as providing an estimation of the factor of safety and the probable failure
mechanism, but engineers should also appreciate the limitations of each method when
assessing the results of their analyses.





















Chapter 3: Two-dimensional Slope Stability Analysis by Limit Equilibrium and Strength Reduction Methods

- 50 -

Table 3.1 Factors of safety by LEM and SRM
case
c
(kPa)

()
FOS
(LEM)
FOS
(SRM1,
non-
associated)
FOS
(SRM2,
associated)
FOS
difference
with LEM
(SRM1, %)
FOS
difference
with LEM
(SRM2, %)
FOS
difference
between
SRM1 and
SRM2 (%)
1 2 5 0.25 0.25 0.26 0 4.0 4.0
2 2 15 0.50 0.51 0.52 2 4.0 2.0
3 2 25 0.74 0.77 0.78 4.0 5.4 1.3
4 2 35 1.01 1.07 1.07 5.9 5.9 0
5 2 45 1.35 1.42 1.44 5.2 6.7 1.4
6 5 5 0.41 0.43 0.43 4.9 4.9 0
7 5 15 0.70 0.73 0.73 4.3 4.3 0
8 5 25 0.98 1.03 1.03 5.1 5.1 0
9 5 35 1.28 1.34 1.35 4.7 5.5 0.7
10 5 45 1.65 1.68 1.74 1.8 5.5 3.6
11 10 5 0.65 0.69 0.69 6.2 6.2 0
12 10 15 0.98 1.04 1.04 6.1 6.1 0
13 10 25 1.30 1.36 1.37 4.6 5.4 0.7
14 10 35 1.63 1.69 1.71 3.7 4.9 1.2
15 10 45 2.04 2.05 2.15 0.5 5.4 4.9
16 20 5 1.06 1.20 1.20 13.2 13.2 0
17 20 15 1.48 1.59 1.59 7.4 7.4 0
18 20 25 1.85 1.95 1.96 5.4 5.9 0.5
19 20 35 2.24 2.28 2.35 1.8 4.9 3.1
20 20 45 2.69 2.67 2.83 0.7 5.2 6.0
21 5 0 0.20 0.23 15
22 10 0 0.40 0.45 12.5
23 20 0 0.80 0.91 13.8




Chapter 3: Two-dimensional Slope Stability Analysis by Limit Equilibrium and Strength Reduction Methods

- 51 -

Table 3.2 Soil properties for Figure 3.6
Soil name
Cohesion
(kPa)
Friction angle
(degree)
Density
(kN/m
3
)
Elastic
modulus (MPa)
Poisson ratio
Soil1 20 35 19 14 0.3
Soil2 0 25 19 14 0.3
Soil3 10 35 19 14 0.3


Table 3.3a FOS by SRM from different programs when c=0 and =25 for soft band.
The values in each cell are based on SRM1 and SRM2 respectively. (min. FOS=0.927
from Morgenstern-Price analysis)
Program/FOS 12m domain 20m domain 28m domain
Flac3D 1.03/1.03 1.30/1.28 1.64/1.61
Phase2 0.77/0.85 0.84/1.06 0.87/1.37
Plaxis 0.82/0.94 0.85/0.97 0.86/0.97
Flac2D No solution No solution No solution



Table 3.3b FOS by SRM from different programs when =0 and c=10 kPa for soft
band. The values in each cell are based on SRM1 and SRM2 respectively. (min.
FOS=1.03 from the Morgenstern-Price analysis)
Program/FOS 28m domain
Flac3D 1.06/1.06
Phase2 0.99/1.0
Plaxis 1.0/1.03
Flac2D No solution









Chapter 3: Two-dimensional Slope Stability Analysis by Limit Equilibrium and Strength Reduction Methods

- 52 -

Table 3.4 FOS with non-associated flow rule for 12m domain
Element number
Tolerance (stress
analysis)
Maximum
number of
iterations
FOS
1500 0.001 100 0.80
2000 0.001 100 No result
2000 0.003 100 No result
2000 0.004 100 No result
2000 0.005 100 No result
2000 0.008 100 0.81
2000 0.01 100 0.82
2000 0.001 500 0.74
2000 0.003 500 0.77
2000 0.004 500 0.77
2000 0.005 500 0.79
3000 0.001 100 No result
3000 0.003 100 0.79
3000 0.004 100 0.8
3000 0.005 100 0.8
3000 0.01 100 0.84
3000 0.001 500 0.77


Table 3.5 FOS with associated flow rule for 12m domain
Element number
Tolerance (stress
analysis)
Maximum
number of
iterations
FOS
1000 0.001 100 1.03
1200 0.001 100 1.0
1500 0.001 100 No result
1500 0.003 100 No result
1500 0.004 100 1
1500 0.005 100 1.39
1500 0.01 100 2.09
1500 0.001 500 0.86
1500 0.003 500 0.98
3000 0.001 100 No result
3000 0.003 100 No result
3000 0.004 100 No result
3000 0.005 100 No result
3000 0.01 100 No result
3000 0.001 500 0.85
3000 0.003 500 0.89
3000 0.004 500 0.9
3000 0.005 500 1.5
3000 0.01 500 2.09

Chapter 3: Two-dimensional Slope Stability Analysis by Limit Equilibrium and Strength Reduction Methods

- 53 -

Table 3.6 Comparison of factor of safety for Figure 3.14
case
c
(kPa)

()
Density
(kN/m
3
)
Poisson
ratio
Elastic
modulus
of soil1
(kN/m
2
)
Elastic
modulus
of soil2
(kN/m
2
)
FOS
(LEM)
FOS
(SRM1)
FOS
(SRM2)
1 10 15 20 0.3 1.4e5 1.4e2 0.9826 0.954 0.983
2 10 15 20 0.3 1.4e2 1.4e5 0.9826 0.966 0.989
3 10 25 20 0.3 1.4e5 1.4e2 1.2951 1.235 1.297
4 10 25 20 0.3 1.4e2 1.4e5 1.2951 1.259 1.320

Table 3.7 Comparison of factor of safety for Figure 3.15
case
c
(kPa)

()
Density
(kN/m
3
)
Poisson
ratio
Elastic
modulus
of soil1
(kN/m
2
)
Elastic
modulus
of soil2
(kN/m
2
)
FOS
(LEM)
FOS
(SRM1)
FOS
(SRM2)
1 10 15 20 0.3 1.4e5 1.4e2 0.982 0.964 0.985
2 10 15 20 0.3 1.4e2 1.4e5 0.982 0.963 0.976
3 10 25 20 0.3 1.4e5 1.4e2 1.295 1.288 1.304
4 10 25 20 0.3 1.4e2 1.4e5 1.295 1.282 1.273

Table 3.8 Comparison of factor of safety for Figure 3.16
case
c
(kPa)

()
Density
(kN/m
3
)
Poisson
ratio
Elastic
modulus
of soil1
(kN/m
2
)
Elastic
modulus
of soil2
(kN/m
2
)
FOS
(LEM)
FOS
(SRM1)
FOS
(SRM2)
1 10 15 20 0.3 1.4e5 1.4e2 0.9826 0.951 0.980
2 10 15 20 0.3 1.4e2 1.4e5 0.9826 0.975 0.980
3 10 25 20 0.3 1.4e5 1.4e2 1.2951 1.240 1.292
4 10 25 20 0.3 1.4e2 1.4e5 1.2951 1.240 1.275

Table 3.9 Comparison of factor of safety for Figure 3.17
case
c
(kPa)

()
Density
(kN/m
3
)

Elastic
modulus
of soil1
(kN/m
2
)
Elastic
modulus
of soil2
(kN/m
2
)
Elastic
modulus
of soil3
(kN/m
2
)
FOS
(LEM)
FOS
(SRM1)
FOS
(SRM2)
1 10 15 20 0.3 1.4e5 1.4e2 1.4e5 0.940 0.932 0.951
2 10 15 20 0.3 1.4e2 1.4e5 1.4e2 0.940 0.903 0.947
3 10 25 20 0.3 1.4e5 1.4e2 1.4e5 1.261 1.210 1.275
4 10 25 20 0.3 1.4e2 1.4e5 1.4e2 1.261 1.211 1.268



Chapter 3: Two-dimensional Slope Stability Analysis by Limit Equilibrium and Strength Reduction Methods

- 54 -


4m
4m 6m 10m
10m
20m
6m

Figure 3.1 Discretization of a simple slope model

0
1
2
3
4
5
6
7
8
9
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
x(m)
y
(
m
)
ground profile
limit equilibrium
SRM1
SRM2
0
1
2
3
4
5
6
7
8
9
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
x(m)
y
(
m
)
ground profile
limit equilibrium
SRM1
SRM2

(a) c=2kPa, =5 (b) c=2kPa, =15
0
1
2
3
4
5
6
7
8
9
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
x(m)
y
(
m
)
ground profile
limit equilibrium
SRM1
SRM2
0
1
2
3
4
5
6
7
8
9
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
x(m)
y
(
m
)
ground profile
limit equilibrium
SRM1
SRM2

(c) c=2kPa, =25 (d) c=2kPa, =35
0
1
2
3
4
5
6
7
8
9
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
x(m)
y
(
m
)
ground profile
limit equilibrium
SRM1
SRM2

(e) c=2kPa, =45
Figure 3.2 Slip surface comparison with increasing friction angle (c=2kPa)
Chapter 3: Two-dimensional Slope Stability Analysis by Limit Equilibrium and Strength Reduction Methods

- 55 -


0
1
2
3
4
5
6
7
8
9
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
x(m)
y
(
m
)
ground profile
limit equilibrium
SRM1
SRM2
0
1
2
3
4
5
6
7
8
9
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
x(m)
y
(
m
)
ground profile
limit equilibrium
SRM1
SRM2

(a) c=2kPa, =5 (b) c=5kPa, =5
0
1
2
3
4
5
6
7
8
9
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
x(m)
y
(
m
)
ground profile
limit equilibrium
SRM1
SRM2
0
1
2
3
4
5
6
7
8
9
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
x(m)
y
(
m
)
ground profile
limit equilibrium
SRM1
SRM2

(c) c=10kPa, =5 (d) c=20kPa, =5
Figure 3.3 Slip surface comparison with increasing cohesion (=5)


0
1
2
3
4
5
6
7
8
9
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
x(m)
y
(
m
)
ground profile
limit equilibrium
SRM1
SRM2
0
1
2
3
4
5
6
7
8
9
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
x(m)
y
(
m
)
ground profile
limit equilibrium
SRM1
SRM2

(a) c=2kPa, =35 (b) c=5kPa, =35
0
1
2
3
4
5
6
7
8
9
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
x(m)
y
(
m
)
ground profile
limit equilibrium
SRM1
SRM2
0
1
2
3
4
5
6
7
8
9
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
x(m)
y
(
m
)
ground profile
limit equilibrium
SRM1
SRM2

(a) c=10kPa, =35 (b) c=20kPa, =35
Figure 3.4 Slip surface comparison with increasing cohesion (=35)


Chapter 3: Two-dimensional Slope Stability Analysis by Limit Equilibrium and Strength Reduction Methods

- 56 -




0
2
4
6
8
10
12
14
16
18
20
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42
x(m)
y
(
m
)
ground profile
limit equilibrium
SRM

0
2
4
6
8
10
12
14
16
18
20
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42
x(m)
y
(
m
)
ground profile
limit equilibrium
SRM

(a) c=5kPa, =0 (b) c=20kPa, =0
Figure 3.5 Slip surface comparison with increasing cohesion (=0)




x
y
0,0
0,5
5,5
8,8
20,15
5,4.5
8,7.5
8,7.1
28,15
28,10
28,9.5
28,0
Soil1
Soil2
Soil3

Figure 3.6 A slope with a thin soft band






Chapter 3: Two-dimensional Slope Stability Analysis by Limit Equilibrium and Strength Reduction Methods

- 57 -






(a) numerical simulation model 1

(b) numerical simulation model 2 (c) numerical simulation model 3
28m
20m 12m

Figure 3.7 Mesh plot of the three numerical models with a soft band






Chapter 3: Two-dimensional Slope Stability Analysis by Limit Equilibrium and Strength Reduction Methods

- 58 -




0 5 10 15 20 25
0
2
4
6
8
10
12
14
soil1
soil2
soil3

(a) Critical solution from LEM when soft band is frictional material (FOS=0.927)


(b) Critical solution from SRM for 12m width domain
Figure 3.8 Critical failure surfaces from LEM and SRM for frictional soft band problem




Chapter 3: Two-dimensional Slope Stability Analysis by Limit Equilibrium and Strength Reduction Methods

- 59 -


(a) Critical failure surface from LEM when the bottom soil layer is weak (FOS=1.29)


(b) Critical failure surface from SRM2 and 20m domain (FOS=1.33)
Figure 3.9 Critical solutions from LEM and SRM when the bottom soil layer is weak

0,0
0,10 10,10
16,16
26,21
46,21
46,0
x
y
1V
2H
1V
1H
Soil property:
Cohesion=10kPa
Friction angle=30degree
Density=20kN/m
3
Elastic modulus=14MPa
Poisson ratio=0.3

Figure 3.10 Slope geometry and soil property under study

Chapter 3: Two-dimensional Slope Stability Analysis by Limit Equilibrium and Strength Reduction Methods

- 60 -






Figure 3.11 Result derived by SRM



FOS=1.3848
FOS=1.4019
FOS=1.3830
FOS=1.4063
FOS=1.4480
Figure 3.12 Global and local minima by LEM





Chapter 3: Two-dimensional Slope Stability Analysis by Limit Equilibrium and Strength Reduction Methods

- 61 -
0 5 10 15 20 25 30 35 40 45 50 55
0
5
10
15
20
25
30
silt

(a) FOS=1.33 (global minimum)
0 5 10 15 20 25 30 35 40 45 50 55
0
5
10
15
20
25
30
silt

(b) FOS=1.375 (local minimum)
0 5 10 15 20 25 30 35 40 45 50 55
0
5
10
15
20
25
30
silt

(c) FOS=1.415 (local minimum)
Chapter 3: Two-dimensional Slope Stability Analysis by Limit Equilibrium and Strength Reduction Methods

- 62 -
0 5 10 15 20 25 30 35 40 45 50 55
0
5
10
15
20
25
30
silt

(d) FOS=1.40 (local minimum)
0 5 10 15 20 25 30 35 40 45 50 55
0
5
10
15
20
25
30
silt

(e) FOS=1.383 (local minimum)

(f) FOS=1.327 from SRM
Figure 3.13 Local minima from LEM and critical solution from SRM for a simple slope

Chapter 3: Two-dimensional Slope Stability Analysis by Limit Equilibrium and Strength Reduction Methods

- 63 -

0,0
0,5
5,5
21,0
21,8
8,8
11,11 21,11
Soil1
Soil2
X
Y

Figure 3.14 Slope geometry with two soils and a horizontal boundary

0,0
0,5
5,5
21,0
8,0
8,8
11,11 21,11
Soil1 Soil2
X
Y

Figure 3.15 Slope geometry with two soils and a vertical boundary

0,5
5,5
21,0
8,8
11,11 21,11
Soil1
Soil2
X
Y
11,10
5,4
0,0
0,4
21,8

Figure 3.16 Slope geometry with two soils and an inclined boundary

Chapter 3: Two-dimensional Slope Stability Analysis by Limit Equilibrium and Strength Reduction Methods

- 64 -


0,5
5,5
29,0
11,9
11,11
29,17
Soil1
Soil2
X
Y
11,10
5,4
0,0
0,4
29,6
0,3
5,3
29,5
Soil3

Figure 3.17 Slope geometry with three soils and an inclined boundary



LEM
SRM

Figure 3.18 Critical slip surface of case 2 in Figure 3.17 based on SRM and LEM
Chapter 4:Three-dimensional Slope Failure Analysis by the Strength Reduction and Limit Equilibrium Methods

- 65 -

CHAPTER 4: THREE-DIMENSIONAL SLOPE FAILURE
ANALYSIS BY THE STRENGTH REDUCTION AND LIMIT
EQUILIBRIUM METHODS

4.1 Introduction
So far, there have only been limited investigations of 3D SRM and 3D LEM methods,
and there is also a lack of comparison between these two methods. In this chapter, an
extensive study on 3D slope failure based on these two methods is conducted for several
typical non-reinforced slopes with obvious 3D effects, and some interesting and useful
findings not present in the corresponding two-dimensional analysis are obtained. The
suitability of these two methods under different conditions is investigated and
precautions to apply these methods are suggested. The 3D LEM formulation adopted in
the present study is based on that by Cheng et al. (2005) and Cheng and Yip (2007),
while the associated flow rule is used for all the SRM studies in this chapter.

4.2 Stability analysis for a vertical cut
Consider the vertical cut slope shown in Figure 4.1. The height of the slope is 5 m while
the cohesive strength (c), friction angle () and unit weight of the soil are 24.5 kPa, 20
and 17.64 kN/m
3
respectively. The FOS determined by the 3D SRM using FLAC3D is
1.51 while the critical failure surface is shown in Figure 4.1. It appears that the slip
surface is composed of two parts. The first failure mode is a 2D failure along the x-
direction while the other failure mode is a 2D failure along the y-direction. The two
parts are perpendicular and intersect with each other. To obtain more insight into the
failure mode, another model which is unconstrained on only one vertical plane is
Chapter 4:Three-dimensional Slope Failure Analysis by the Strength Reduction and Limit Equilibrium Methods

- 66 -
developed and the FOS determined by the SRM is 1.55 (Figure 4.2). It can be seen that
the FOS for a vertical cut that is unconstrained on two vertical planes is only slightly
smaller than that for the vertical cut that is unconstrained on only one vertical plane.
The FOS is practically the same in these two cases, which means that the three-
dimensional effect is small in this example.

The sliding angle in Figure 4.1 determined by the SRM is equal to 0 or 90. It is
suspected that the sliding direction should be at a direction of 45 in the x-y plane for
this vertical cut, so wedge failures as shown in Figures 4.3 and 4.4 are considered in the
analysis. Two different sliding directions are proposed for the vertical cut that is
unconstrained on two vertical planes. Firstly, a wedge block that fails along a direction
of 45 in the x-y plane is considered (Figure 4.3). Secondly, the wedge is supposed to
fail along zero degree direction (Figure 4.4). In Figure 4.3, the weight of the three-
dimensional sliding wedge block ABCD can be expressed as
3
cot
W
2 3
h
= (4.1)
where is the density of the soil, h is the height of the vertical cut, is the dip angle of
the sliding plane, and W is the weight of the wedge block. The FOS can be expressed as


sin
1 cot cot tan cos
FOS
2 2
W
h c W + +
= (4.2)
Introducing eq. (4.1) into (4.2) gives the FOS equal to:


2 sin
6 tan 2 cos tan
FOS
h
c h h + +
= (4.3)
The minimum value of FOS can be obtained by differentiating eq.(4.3), which gives:
0 2 cos 6 2 cos tan tan h = + + c h (4.4)
From eq.(4.4), we can get
Chapter 4:Three-dimensional Slope Failure Analysis by the Strength Reduction and Limit Equilibrium Methods

- 67 -
c h
h
+

=
6 tan
tan
2 cos


(4.5)
Put =17.64 kN/m
3
, h=5m, =20 and c=24.5 kPa into eqs. (4.5) and (4.3), gives =
50.16 while the minimum FOS is 2.0 which is much greater than 1.51 or 1.55 for the
3D SRM.

In Figure 4.4, the sliding angle is supposed to be zero, so it can be simplified to a two-
dimensional sliding wedge block. The weight of the two-dimensional sliding wedge
block OEF can be expressed as
2
cot
W
2
h
= (4.6)
Then the FOS can be expressed as


sin
sin / tan cos
FOS
W
h c W +
= (4.7)
Introducing eq.(4.6) into (4.7),


2 sin
4 tan 2 cos tan
FOS
h
c h h + +
= (4.8)
Differentiating eq.(4.8) and setting it equal to zero gives
0 2 cos 4 2 cos tan tan h = + + c h (4.9)
From eq. (4.9), we can get
c h
h
+

=
4 tan
tan
2 cos


(4.10)
Put =17.64 kN/m
3
, h=5m, =20 and c=24.5kPa into eqs. (4.10) and (4.8), a
minimum FOS 1.43 is obtained at =52.14, which is very close to the value of 1.5
determined by the 3D SRM. While the real failure mechanism may not be a wedge
failure mechanism, the FOS obtained from the wedge failure mechanism which neglects
the inter-column shear forces will provide an estimate of the factor of safety.
Chapter 4:Three-dimensional Slope Failure Analysis by the Strength Reduction and Limit Equilibrium Methods

- 68 -
From the above analysis, it can be seen that it is actually more difficult for the vertical
cut to slide along a 45 degree direction, and this can also be seen from the relationship
between the FOS and dip angle of the sliding plane shown in Figure 4.5. Hence, the
critical failure surface obtained by the 3D SRM, which is composed of two parts along
the x and y directions, should be the reasonable failure mode. As a further check, the
vertically cut slope is also analyzed by the 3D LEM method in the program Slope3D by
Cheng et al. (2005) and Cheng and Yip (2007), and a FOS 1.47 with a 14 sliding
direction is obtained from the optimization analysis by the 3D J anbu method, giving
results that are also close to those obtained by the SRM. Though the sliding angle from
the LEM is not zero as it is in the SRM, this result is closer to zero than 45.

Although the FOS for the vertical cut that is unconstrained on two adjoining vertical
planes is nearly the same as the one for the vertical cut that is unconstrained on only one
vertical plane, there are significant differences when there is local loading on the top of
the slope and this issue will be examined in detail in the later section about the effect of
curvature on slope stability.

From Figure 4.1, it can also be seen that the shear strain at the top of the vertical cut is
larger at the location where the two slip surfaces from the x and y directions intersect.
At the slope toe, the shear strain is also larger near the intersection of the two
unconstrained planes, and becomes smaller at increasing distances away from the
intersection. To consider the boundary effect, a very large model is developed (Figure
4.6). The FOS for this model is 1.56, which is nearly the same as that for the small
model (1.51), and the small difference is due to the element size effect. The slip surface
is still composed of two parts, but in contrast to the small model, the slip surface does
not extend to the solution boundary in the large model. This shows that even though the
Chapter 4:Three-dimensional Slope Failure Analysis by the Strength Reduction and Limit Equilibrium Methods

- 69 -
FOS for the vertical cut with two unconstrained planes is nearly the same as that with
one unconstrained plane, it is still slightly weaker near to the intersection of the two slip
surfaces.

In the previous SRM analysis, the finite difference method in the Flac3D software is
used. In addition, the finite element program Midas is also used for the 3D SRM in the
present study. In the finite element code, the non-convergence criterion occurs when
convergence is not achieved after a specified number of iterations. It has been
demonstrated in chapter 3 that the choice of the tolerance in the solution to the nonlinear
equations and the number of iterations allowed (in the 2D SRM) can be critical in the
analysis. There are no simple guidelines for choosing optimum values for general cases,
though the default setting in commercial programs appears to be reasonable except for
isolated cases. As shown in Table 4.1, when the convergence criterion and the
maximum number of iterations vary, the FOS for the vertically cut slope varies from
0.9375 to 2.3125. Since the range of the solution is very large, it will not be easy to
choose the proper convergence criterion and the maximum number of iterations in the
3D SRM analysis. It has been shown that the reliance on the computer program default
setting may not be adequate for the 2D SRM, and this is also becomes a critical issue
for the 3D SRM. Great care and judgment, as well as some trial and error, may be
required before a reasonable result can be obtained, which is a major disadvantage in
the application of the 3D SRM to routine analysis and design. It is also suggested that
results from the 3D LEM are determined for comparison with the results from the 3D
SRM.

For the case in Figure 4.1, the displacement criterion (Naylor, 1982; Donald and Giam,
1988; Griffiths and Lane, 1999) has also been tried in which the slope failure is
Chapter 4:Three-dimensional Slope Failure Analysis by the Strength Reduction and Limit Equilibrium Methods

- 70 -
determined by a sudden change in the displacement instead of relying on the computer
program default. The displacement at the corner of the slope crest is shown in Figure
4.7, when the strength reduction factor increases from 1.4 to 1.5 the displacement
changes suddenly, so we can assume that the FOS is around 1.5. This result agrees well
with the results from the finite difference and limit equilibrium methods. From Figure
4.8, it can be seen that when the strength reduction factor increases from 1.4 to 1.5, the
shear strain also changes suddenly. Some other examples have also been tried, and it is
found that this criterion appears to be a more consistent criterion than the computer
programs default when the results are compared with the 3D LEM. It is, however, not
easy to define a precise factor of safety to better than a few percent accuracy from the
criterion of a sudden change in the displacement.

4.3 Stability analysis for a vertical cut with a weak layer
The second example, which has also been analyzed by Huang and Tsai (2000), is a
vertically cut slope with a 0.5 m thick planar weak layer (Figure 4.9). The FOS obtained
by the 3D SRM model as shown in Figure 4.10a is 0.58 and the slip surface is along the
weak layer (Figure 4.10c). Since the slope fails along the weak layer (c=0, =10) and
the weak layer is inclined at a 35 angle, the FOS can be obtained by a simple
computation as tan10/tan35=0.2518. The FOS obtained by 3D J anbu is 0.2539, which
agrees well with the expected result from basic soil mechanics. The FOS obtained by
the 3D SRM is 0.58, which is much larger than the reasonable value of 0.25, and the
model in Figure 4.10a has been carefully checked. When the vertically cut slope is
about to fail, a large shear strain will be mobilized at the slope toe. As shown in Figure
4.10c, shear strain is localized along the soft band while the bottom part is still stable.
The restraint by the bottom elements on the inclined elements along the soft band will
tend to make failure more difficult and hence a higher FOS will be obtained. Actually,
Chapter 4:Three-dimensional Slope Failure Analysis by the Strength Reduction and Limit Equilibrium Methods

- 71 -
the requirement of continuity from the finite difference or finite element methods is the
major cause for this problem if there is major variation of shear strain across a thin soft
band. This phenomenon can be further illustrated by the model shown in Figure 4.11. If
the element size at the slope toe is very small, a FOS of 0.40 is obtained for this case.
This result is better than the first model, but it is still much larger than the expected
value. A third model as shown in Figure 4.12 is developed, and the weak layer is raised
slightly (by 0.12 m) while the thickness is still 0.5 m. The soft band is still modeled by 4
layers of inclined elements, and only the top layer of elements is not connected to the
bottom which may generate the resistance to failure. Even though this model is
practically the same as the previous model, a continuous shear band can now formed
easily in the top layer of the soft band elements, and a FOS 0.25 is obtained which is the
expected value. The generation of a highly localized shear failure zone similar to that of
a soft band needs great care in the 3D SRM mesh design. For highly complicated 3D
problems, it is not easy to generate a good mesh automatically without any trial and
error.

In this soft band example, the cohesion of the soft band is zero. In order to investigate
the results when the friction angle of the soft band is very small, another model is
developed in which cohesion is equal to 10 kPa and the friction angle is zero. The
factors of safety obtained by the first model (Figure 4.10a), the second model (Figure
4.11a) and the third model (Figure 4.12a) are 0.75, 0.56 and 0.47 respectively. The FOS
for this problem is given by eq. (4.8) as 0.4826. The third numerical model gives a FOS
0.47 which is very close to the LEM result of 0.4826. The first model is poor as the FOS
obtained by this model is much larger than the expected value. The FOS obtained by the
second model is larger than the expected value, but the situation is better than that in the
previous example where the cohesion strength is 0, so it seems that when the friction is
Chapter 4:Three-dimensional Slope Failure Analysis by the Strength Reduction and Limit Equilibrium Methods

- 72 -
zero for the soft band, the influence of the mesh design is slightly smaller than that in
the previous case.

4.4 Stability analysis for a slope with transverse earthquake load
Cheng and Yip (2007) have proposed a new 3D asymmetrical LEM formulation. Under
this new formulation, it is found that when a transverse earthquake load is present,
Huang and Tsais method (2000, 2002) has difficulty in converging because the sliding
direction varies between columns. On the other hand, Cheng and Yips method adopts a
single sliding direction and does not encounter such convergence problems. Cheng and
Yip (2007) have considered an earthquake coefficient of up to 0.5, but a higher
earthquake coefficient which may not be realistic will also be studied in this section.
The slope model under consideration is shown in Figure 4.13. The slope height is 6 m,
the slope angle is 45, the soil unit weight is 20.0 kN/m
3
, the cohesion is 5 kPa and the
friction angle is 35.

The results of the transverse earthquake load analyses are shown in Tables 4.2 and 4.3.
In Table 4.2, the earthquake coefficient in the y-direction (transverse) is 0.5, while the
earthquake coefficient load in the x-direction varies from 0.1 to 1.0. Although the FOS
obtained by the SRM are larger than those by the LEM, the differences are not great.
With an increase in the earthquake load in the x-direction, the FOS becomes smaller,
and this is also a very reasonable result. For the SRM analysis, the average of the
sliding direction is used in Tables 4.2 and 4.3. Although with increasing earthquake
load in the x-direction the sliding angle obtained by the SRM also gets smaller, the
angle is smaller than that obtained by the LEM. This difference may be caused by the
different mechanisms in the LEM and SRM. In LEM analysis, the sliding mass is
assumed to be a rigid block and it moves in the same direction at initiation of slope
Chapter 4:Three-dimensional Slope Failure Analysis by the Strength Reduction and Limit Equilibrium Methods

- 73 -
failure. In the SRM analysis, the stress-strain relation is used and the velocity and
displacement for each element of the sliding mass are different.

In Table 4.3, the earthquake load in the x-direction is 30% of the soil weight, while the
earthquake load in the y-direction varies from 1% to 200%. When the earthquake load
in the y-direction varies from 1% to 200%, the two results obtained by the SRM and
LEM agree well even though the earthquake load is very large and the FOS gets smaller
with the increase in earthquake load. It is also reasonable that the FOS gets smaller
slowly but the sliding direction gets larger drastically as the earthquake load increases in
the y-direction.

4.5 Failure mode due to self weight for a simple infinite slope
Consider a simple slope with a height 6 m and slope angle 45. The unit weight, friction
angle and cohesion of the soil are equal to 20.0 kN/m
3
, 10 and 15 kPa. For a slope with
regular geometry and uniform soil properties, the failure mechanism from the 3D SRM
is practically a 2D failure as shown in Figure 4.14a, which is as expected for an infinite
slope. The minimum 2D and 3D factors of safety based on a spherical search for the
Spencer method are 1.07 and 1.12, respectively. The distinct 3D failure mode is actually
the limitation of the spherical failure surface for the present problem, and this can be
illustrated by the adoption of an ellipsoidal or NURBS search (able to model an almost
2D failure). This search gives a more realistic failure mechanism, close to a 2D failure,
with a minimum factor of safety very close to 1.07. It is hence important to choose a
proper failure shape for locating the critical 3D failure surface in the 3D LEM.

Chapter 4:Three-dimensional Slope Failure Analysis by the Strength Reduction and Limit Equilibrium Methods

- 74 -
To investigate the importance of heterogeneity, it is introduced into this simple slope
with the SRM analysis by defining a Weibull distribution (1951) for the cohesive
strength given by eq.(4.11):
m
c
c
m
e
c
c
c
m
c
) (
1
0 0
0
) ( ) (

= (4.11)
where c
0
is a scale parameter and is related to the average value of the soil cohesive
strength. The parameter m defines the shape of the density function which controls the
degree of material homogeneity, and is referred as the homogeneity index. If m is larger,
it means the soil parameters are more homogenous, while if m gets smaller, it means the
soil parameters are more heterogeneous. The average value of the cohesive strength is
equal to ) / 1 1 ( c
0
m + . The results for different m values are shown in Figure 4.14,
where the average value of cohesion is 15 kPa for both Figures 4.14b and 4.14c. It can
be seen that the introduction of strength heterogeneity into the SRM analysis cannot
generate a distinct 3D failure mechanism. For this case, heterogeneity by Weibull
distribution is not sufficient to generate a distinct 3D failure. If an extreme
heterogeneity is introduced by 3D shape soft band, a 3D failure mode may occur easily.

4.6 Stability analysis for a locally loaded slope
To generate a distinct 3D failure due to the self weight of soil, a much greater spatial
difference in the soil parameters is required for the simple slope in Figure 4.14. On the
other hand, if the failure is induced by external load, a distinct 3D failure will be easily
formed. Consider the slope with a rectangular area of vertical loading as shown in
Figure 4.15. The loading width B is 2 m while the edge of the loading is 1 m away from
the crest of the slope. The slope geometry and the soil properties are shown in Figure
4.15. Results of the analysis are shown in Table 4.4, for a loading q of 100 kPa and
where the ratio of the loading length L to loading width B varies from 0 to 10. In the 3D
Chapter 4:Three-dimensional Slope Failure Analysis by the Strength Reduction and Limit Equilibrium Methods

- 75 -
SRM analysis, when the loading length is 2 m, 4 m, 8 m or 12 m (or L/B=1, 2, 4 and 6),
the length of model chosen is 20 m. When the loading length is 16 m or 20 m (viz.
L/B=8 and 10), the model length is chosen to be 30 m and the results obtained by the
3D SRM and LEM agree well with the expected pattern of decreasing FOS with
increasing loading length. The slip surfaces obtained by the SRM are shown in Figure
4.16. When L/B=1, the slip surface is basically a two-dimensional failure similar to that
in Figure 4.14a. Although greater shear strain is mobilized around the loading, the
loading is not large enough to mobilize the shear strain to form a 3D slip surface. That
means, the effect of the self weight of soil controls the failure so that a 2D slip surface
will be obtained. If L/B is increased so that the external load becomes more significant,
then the failure will be controlled by the external load. When L/B=2, a nearly three-
dimensional slip surface is formed by the SRM. When L/B4, a very clear 3D failure
can be mobilized. If L is very long, a 2D failure will appear again. In conclusion, when
L is very small (close to B), the small external load is shared by a much greater soil
mass so that the effect of the external load is small and not effective in generating a 3D
failure mode. When L is very large, the problem is clearly a 2D problem, while for the
intermediate cases, the external load can generate a distinct 3D failure. If the external
load is very high, then a small L can also generate a distinct 3D failure mode for a
simple slope. Since the self weight is not effective in generating a true 3D failure in the
SRM for a simple slope, the magnitude of the external load superimposed onto the self
weight becomes critical in determining the actual failure mode.

To investigate the boundary effect, different model lengths have been tried for L/B=1
and L/B=4 and the results are shown in Table 4.5. For L/B=1, when the model length
varies from 10 m to 30 m, the FOS increases from 1.62 to 1.74. For L/B=4, when the
model length varies from 10 m to 20 m, the FOS increases from 1.27 to 1.41 and
Chapter 4:Three-dimensional Slope Failure Analysis by the Strength Reduction and Limit Equilibrium Methods

- 76 -
remains constant with further increase in the model length. It can be seen that for
L/B=4, a 20 m model length is good enough while for L/B=1, when the model length
increases from 20 m to 30 m, the FOS still increases slightly. This seems to be contrary
to what one would intuitively expect in numerical analysis. If the differences in the
failure mechanism for these two cases are considered, the results can be explained
easily.

For L/B=4, the failure mechanism is a distinct 3D surface. In Figure 4.17, it can be seen
that when the model length is 14 m, the slip surface is basically 2D as the boundary is
close to the external load which inhibits the development of a distinct 3D failure. When
the model length increases to 20 m, the boundary effect disappears and a clear 3D slip
surface is formed so that the FOS will remain constant with any further increase in the
model length. That means, a finite boundary is sufficient for a true 3D analysis.

For failure controlled by the self weight instead of the external load, the situation will
be different. For L/B=1, the true slip surface is 2D because the applied load is small
compared with the self weight of the soil, and the FOS for the 2D failure is smaller than
for a 3D failure. When there is a small applied load, either a distinct 3D failure
mechanism will operate, or an approximately 2D failure with the applied load shared
among the whole failure mass will be formed. The actual failure mechanism will be
controlled by the mechanism with a smaller FOS, and an approximately 2D failure is
still possible for a slope with a small patchy load. When the boundary effect is
considered, the failure mechanism is an important factor. The failure mechanism is
controlled by both the loaded length and the loading intensity. As shown in Figure 4.18,
for L/B=4, when the loading is 20 kPa, the true slip surface is still a 2D failure. When
the loading increases to 100 kPa, a distinct three-dimensional slip surface is formed; and
Chapter 4:Three-dimensional Slope Failure Analysis by the Strength Reduction and Limit Equilibrium Methods

- 77 -
when the loading further increases to 300 kPa, the 3D slip surface is more distinct with
a smaller sliding mass volume with increasing external load. The failure mechanism
hence varies with the loading length and loading intensity. For a locally loaded slope,
the critical failure may range from 2D to 3D, which is not known before the analysis.
Based on various trials on many examples, the following procedure is proposed for the
assessment of a locally loaded slope which is applicable for both the SRM and the LEM
analyses.

If the slope is short in length, the entire slope should be modeled according to its true
boundary conditions irrespective of its failure mechanism. If the real slope is very long,
the analysis can be conducted as follows.

(1) A certain model length is chosen by experience or by intuition. SRM analysis is then
conducted. If the failure mode is 3D, this will be the solution to the problem. If the
failure mode is 2D, we should go to the next step for a more detailed analysis (based on
the studies as shown above).

(2) Conduct SRM analysis with a longer model length. If the failure mode changes from
2D to 3D, this will be the solution of the problem. If the failure mode is still 2D, we
should compare the current FOS value with the one from step 1 above. If the increase in
the FOS is very small, we can take this result as the solution of the problem, otherwise,
further increase the model length and repeat step 2 until a final judgment can be made.

Finally, there is one important point that must be noted. If the slope is very long and the
local loading is small so that the critical failure mode is 2D, when the model length
increases to infinity, the FOS will, theoretically, converge to that of a 2D slope without
Chapter 4:Three-dimensional Slope Failure Analysis by the Strength Reduction and Limit Equilibrium Methods

- 78 -
external load in the SRM. This result would tend to over-estimate slope safety as the
mobilized shear strain around the local loading is slightly greater than the remaining soil
mass even though a distinct 3D failure mode is not formed. A similar situation will also
occur in the 3D LEM as the load is shared by a large volume of soil mass, but a distinct
3D failure can usually be formed more easily by the 3D LEM as compared with the
SRM. In this case, a comparison with the 3D LEM is suggested as a check to the SRM
result. In addition, to decide the length of the model for analysis with 3D SRM, the
boundary effect should be removed and a procedure to determine the appropriate model
length is recommended in the above. From the results as shown in this chapter, 2.5L
seems to be acceptable for common situations, so this model length can be chosen in the
first stage of SRM simulation.

4.7 Curvature effect on the slope stability
In this section, the influence of the curvature on the slope stability is investigated.
Firstly, several idealized concave and convex slope models are developed. The slope
height is 10m and slope angle is 45 with properties of density=19.5kN/m
3
, c=10kPa
and =36. In this analysis, two different groups of slope models are developed. In the
first group, no external loading is applied on the slope. In the second group, there is a
4m width 100 kPa external loading applied on the crest of slope. In each group, five
models are developed (two concave models with 17m and 35m radius at the slope
middle height, two convex models with 17m and 35m radius at the slope middle height
and one plane model with no curvature). The typical slope section is shown in Figure
4.19. The typical geometry of the convex and concave slopes is shown in Figure 4.20.
Since the strength reduction analysis is time consuming in 3D analysis, if the model size
is very large as shown in Figure 4.20, it will take a long time for the analysis, and this
Chapter 4:Three-dimensional Slope Failure Analysis by the Strength Reduction and Limit Equilibrium Methods

- 79 -
problem is particularly serious for soil nailed slope which will be analyzed in next
chapters. To take advantage of symmetry in the present problem to reduce the size of
the problem, only a slice mode is adopted in the 3D analysis which is adequate for the
present problem. The slice model has a thickness of 1.5m at the middle of the slope, and
the typical models are shown in Figure 4.21.

The variation of the safety factors with different curvatures are shown in Tables 4.6 to
4.7. For concave slope, the effect of the curvature is obvious and the FOS increases with
the decrease of radius. For convex slope, it appears that the curvature does not have any
major effect on the factor of safety, as the FOS is very close to that for a plane problem.

The slip surfaces for different curvature with no loading is applied are shown in Figure
4.22. For slope with loading on the top, the slip surface is influenced by the loading
location which is a typically local failure, so the slip surface is not compared for slope
with local loading. The slip surfaces for different curvature are nearly the same. This
means that the curvature has no major effect on the location of the slip surface under
normal curvature (there are some minor effect for extreme curvature).

In the above, several idealized concave and convex slopes are investigated. In practice,
the curvature of the slope is usually not idealized as shown in the previous paragraphs
(Figure 4.20), and intersected slope is a popular alternative for construction. Two
different intersected slope models are thus developed for this study. For the concave
slope model, the intersection angle is 135 and for the convex slope the intersection
angle is 145. The slope height is 6 m and the slope angle is 45.

Chapter 4:Three-dimensional Slope Failure Analysis by the Strength Reduction and Limit Equilibrium Methods

- 80 -
Firstly, no loading is applied on the slope, and the FOS and slip surface for different
curvature are shown in Figure 4.23. The FOS for convex slope is nearly the same as that
for a plane slope, but the FOS for concave slope is larger and this result is similar with
the previous cases that the curvature does not have any effect on the factor of safety for
convex slope. For idealized curved slope shown in the previous paragraphs, the
curvature has no effect on the location of slip surface, but for intersected slope, the slip
surface is shallower at the intersection area for concave slope and it is deeper at the
intersection area for convex slope.

Secondly, rectangular shape vertical loading is applied on the top of the slope. The
loading length is 4m and the loading width B is 2m, while the edge of the loading is 1m
away from the crest of the slope. The FOS and slip surface for different curvature are
shown in Figure 4.24. The FOS for convex slope is also nearly the same as no curvature
slope, but the FOS for the concave slope is larger.

In view of the above analysis, it appears that curvature does not has any noticeable
effect on the factor of safety for convex slope, but most engineers view that convex
slope is less safe. In order to get more insight into this issue, the vertical cut slope
(Figure 4.1), which is actually a steeper intersected slope (slope angle is 90) and the
convexity is more obvious (the intersected angle is 90), is analyzed again in the
presence of external loadings. Firstly, 50kPa loading is applied fully along the edge of
the crest and the result is shown in Figure 4.25. The FOS is still nearly the same as that
for a plane slope. Secondly, 3m long and 3m wide 50kPa loading is applied on the
convex corner and the result is compared with a plane slope (shown in Figure 4.26). The
FOS is now much smaller for convex slope. This means that if a slope is steep, the
Chapter 4:Three-dimensional Slope Failure Analysis by the Strength Reduction and Limit Equilibrium Methods

- 81 -
curvature effect will be obvious when local loading is applied on slope top for a convex
slope.

4.8 Discussion and conclusion
The results from the 3D LEM and SRM agree reasonably well for normal situations
where there is no special geological feature with very high or very low shear strength
parameters. In general, the factors of safety obtained by the SRM are slightly higher
than those from the LEM, which is a result similar to the corresponding 2D analysis in
chapter 3.

There are many results in this study which appear to be unexpected, and a detailed
investigation to confirm their validity has been carried out. For the vertical cut with two
unconstrained vertical planes intersecting at 90, the slip surface is found to be
composed of two 2D failure modes. It is actually easier to fail along the x-direction or
y-direction by a 2D failure mode rather than by a 3D failure mode along the diagonal, as
in Figure 4.3. Although the FOS for the vertical cut with two unconstrained planes is
nearly the same as the one for the vertical cut with only one unconstrained plane, there
are great differences in the capacity to resist external loading which are shown in Figure
4.26.

Based on the present study, it is found that the non-convergence criterion in commercial
SRM programs should be assessed carefully, and engineers should not rely completely
on the default setting in the computer programs. In this chapter, it is also found that a
sudden change of the displacement may sometimes be a better criterion than the default
setting in the computer programs for some cases. However, it is difficult to determine a
precise factor of safety according to the classical concept with this criterion. Without
Chapter 4:Three-dimensional Slope Failure Analysis by the Strength Reduction and Limit Equilibrium Methods

- 82 -
comparisons with the LEM, it is not always easy to accept the results from the SRM, as
the results from the SRM may be sensitive to the default settings of the analysis. In fact,
some very poor results have been obtained from commercial programs using the default
settings. In this respect, it seems that the 3D SRM is adequate for normal problems but
care and judgment are required for those special cases, and more research to establish a
more robust criterion for the ultimate limit state under general condition is required. It is
suspected that a more fundamental change in the convergence criterion and nonlinear
solution method may be required for those special cases if human judgment is not
required for the assessment. On the other hand, 3D LEM is relatively insensitive to the
tolerance and the column division which is an advantage over the SRM.

With recent development in the heuristic global optimization methods and the domain
transformation by Cheng et al. (2007) and Cheng (2007), the presence of a soft band is
not a problem in the LEM. From the SRM analysis of the vertically cut slope with a soft
band, the shear strain in the soft band may be difficult to mobilize if the mesh of the
model is not developed properly. For problems with a soft band or major differences in
the soil parameters between soil layers, the numerical model and results should be
checked carefully, and a comparison with the LEM is suggested.

For an infinite slope with local loading, whether the critical failure surface is 2D or 3D
depends on the magnitude of the external load. It is also interesting to find that the
suitable domain size for analysis is largely controlled by the failure mechanism instead
of the length of the external load. As the failure mechanism may be greatly affected by
the boundary effect, for a proper SRM or LEM analysis, the engineers may either adopt
a large domain size at the expense of computer time, or pursue a trial and error analysis
as discussed previously to determine the proper domain size.
Chapter 4:Three-dimensional Slope Failure Analysis by the Strength Reduction and Limit Equilibrium Methods

- 83 -
The basic knowledge in finite element mesh design for non-linear problem will apply to
the 3D SRM, and the importance of the aspect ratio is well covered by Shukha and
Baker (2003). In the present study, the shapes of the elements in SRM are generally
good so that the aspect ratio is not the reason for some of the poor results. For the
vertical cut slope with weak layer as shown in Figure 4.9, the strange FOS is due to
layout of the mesh instead of the aspect ratio, and more attention to the mesh design is
required for the complicated problem using the SRM analysis.

For concave slope, the effect of the curvature is obvious and the FOS will increase with
the decrease of radius. For convex slope, it appears that the curvature does not have any
noticeable effect on the factor of safety for normal situations, and the FOS is nearly the
same as a plane slope. When local loading is applied on the top of convex slope and if
the slope is steep, the effect of curvature will be obvious and the FOS of convex slope
will get smaller. For the idealized curvature slope, the slip surface for different
curvature is nearly the same. For intersected slope, the slip surface is shallower at the
intersection area for concave slope, and it is deeper at the intersection area for convex
slope.









Chapter 4:Three-dimensional Slope Failure Analysis by the Strength Reduction and Limit Equilibrium Methods

- 84 -

Table 4.1 The dependence of FOS on convergence criterion and iteration number
Convergence criterion Iteration number Factor of safety
stress tolerance=0.03 50 1.0375
stress tolerance=0.03 100 1.0375
stress tolerance=0.03 600 1.0875
stress tolerance=0.01 50 0.9375
displacement tolerance=0.001 1000 1.2875
displacement tolerance=0.005 1000 1.3875
displacement tolerance=0.01 1000 2.3125
displacement tolerance=0.03 1000 2.3125
displacement tolerance=0.03 50 2.3125


Table 4.2 Comparison of SRM and LEM results with various earthquake loads in the x-
direction
Earthquake load SRM LEM
Q
x
Q
y
FOS Sliding direction(degree) FOS Sliding direction(degree)
10 50 1.0 27.6 0.90 42.5
30 50 0.8 21.3 0.69 37.9
50 50 0.64 15.8 0.54 32.4
80 50 0.47 9.4 0.37 27.1
100 50 0.38 6.0 0.28 24.4
Note: Q
x
and Q
y
are the earthquake loads (in % of soil weight) in the x- and y-directions respectively


Table 4.3 Comparison of SRM and LEM results with various earthquake loads in the y-
direction
Earthquake load SRM LEM
Q
x
Q
y
FOS Sliding direction(degree) FOS Sliding direction(degree)
30 1 0.90 0.1 0.81 0.8
30 10 0.89 2.0 0.81 8.1
30 30 0.86 12.8 0.77 24.7
30 50 0.80 21.3 0.69 37.9
30 100 0.62 29.2 0.52 56
30 200 0.42 29.0 0.33 70.7

Table 4.4 Safety factors under the local loading of 100kPa
L/B 0 1 2 4 6 8 10
FOS(SRM) 1.82 1.71 1.60 1.41 1.33 1.30 1.26
FOS(LEM) 1.66 1.55 1.47 1.30 1.16 1.13 1.11


Chapter 4:Three-dimensional Slope Failure Analysis by the Strength Reduction and Limit Equilibrium Methods

- 85 -

Table 4.5 Safety factors with different model lengths, obtained by the SRM
Model length(m) 10 14 20 30
L/B=1 1.62 1.68 1.71 1.74
L/B=4 1.27 1.37 1.41 1.41


Table 4.6 Variation of FOS respect to curvature (no load)
Radius of curvature (m) 17 35 (plane)
Concave(SRM) 1.68 1.57 1.48
Convex(SRM) 1.49 1.48 1.48


Table 4.7 Variation of FOS respect to curvature (with 4m width load 100kPa)
Radius of curvature (m) 17 35 (plane)
Concave(SRM) 1.31 1.24 1.16
Convex(SRM) 1.20 1.17 1.16

















Chapter 4:Three-dimensional Slope Failure Analysis by the Strength Reduction and Limit Equilibrium Methods

- 86 -

Z
Y
X

Figure 4.1 Slip surface of a vertical cut with two unconstrained vertical planes
(FOS=1.51)



Figure 4.2 Slip surface of a vertical slope with only one unconstrained vertical plane
(FOS=1.55)




Chapter 4:Three-dimensional Slope Failure Analysis by the Strength Reduction and Limit Equilibrium Methods

- 87 -

A
B
C
D
h
E
Sliding direction on plane
X
Y
Z
45

Figure 4.3 The geometry of the three-dimensional wedge block

O
E
F
h
W
X
Z

Figure 4.4 The geometry of the two-dimensional wedge block

0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
0 10 20 30 40 50 60 70 80 90
Dip angle (degree)
F
a
c
t
o
r

o
f

s
a
f
e
t
y
three dimensional wedge
two dimensional wedge

Figure 4.5 Variation of FOS with respect to the dip angle of the sliding plane


Chapter 4:Three-dimensional Slope Failure Analysis by the Strength Reduction and Limit Equilibrium Methods

- 88 -



Figure 4.6 A large model for a vertical cut (FOS=1.56)


0
0.005
0.01
0.015
0.02
0.025
0.03
0.035
0.04
1 1.1 1.2 1.3 1.4 1.5 1.6
Factor of safety
D
i
s
p
l
a
c
e
m
e
n
t

(
m
)

Figure 4.7 Displacement versus factor of safety





Chapter 4:Three-dimensional Slope Failure Analysis by the Strength Reduction and Limit Equilibrium Methods

- 89 -




(a) FOS =1.4 (b) FOS =1.5
Figure 4.8 Non-linear analysis results using different strength reduction factors


Weak layer
c=0, =10
35
c=24.5 kPa
=20
=17.64 kN/m
3

5m

Figure 4.9 Geometry of a vertical cut with an inclined weak layer (after Huang and Tsai,
2000)



Chapter 4:Three-dimensional Slope Failure Analysis by the Strength Reduction and Limit Equilibrium Methods

- 90 -





(a) mesh for the whole model (b) enlarged view of mesh at toe


(c) slip surface (FOS=0.58)

Figure 4.10 Model one of the strength reduction analysis for a vertical cut with a weak
layer





Chapter 4:Three-dimensional Slope Failure Analysis by the Strength Reduction and Limit Equilibrium Methods

- 91 -





(a) mesh for the whole model (b) enlarged view of mesh at toe


(c) slip surface (FOS=0.40)

Figure 4.11 Model two of the strength reduction analysis for a vertical cut with a weak
layer





Chapter 4:Three-dimensional Slope Failure Analysis by the Strength Reduction and Limit Equilibrium Methods

- 92 -


(a) mesh for the whole model (b) enlarged view of mesh at toe

(c) slip surface (FOS=0.25)
Figure 4.12 Model three of the strength reduction analysis for a vertical cut with a weak
layer



Figure 4.13 Mesh for a slope with transverse earthquake load

Chapter 4:Three-dimensional Slope Failure Analysis by the Strength Reduction and Limit Equilibrium Methods

- 93 -



(a) homogeneous, m=, FOS=1.17 (b) m=3, FOS=1.14 (c) m=1, FOS=1.04
Figure 4.14 Simple slope stability analysis with varying heterogeneity


45
6m
1m 2m
Cohesion =20kN/m
2

Friction angle =20
Density =20kN/m
3

q

Figure 4.15 The geometry of the slope under local loading



(a) L/B=1 (b) L/B=2 (c) L/B=4
Figure 4.16 The slip surfaces for different loading lengths when B=2m



Chapter 4:Three-dimensional Slope Failure Analysis by the Strength Reduction and Limit Equilibrium Methods

- 94 -


(a) 14 m, FOS=1.37 (b) 20 m, FOS=1.41 (c) 30 m, FOS=1.41
Figure 4.17 The slip surfaces for different model lengths when L/B=4



(a) 20 kPa, FOS=1.74 (b) 100 kPa, FOS=1.41 (c) 300 kPa, FOS=0.89
Figure 4.18 The slip surfaces for different local loadings when L/B=4 and model
length=20m



10m
4m
q=100kPa
45

Figure 4.19 The geometry of the slope section



Chapter 4:Three-dimensional Slope Failure Analysis by the Strength Reduction and Limit Equilibrium Methods

- 95 -



(a) convex (b) concave
Figure 4.20 Typical geometry of convex and concave slope



(a) convex (b) concave (c) plane
Figure 4.21 Mesh plot and slip surface of concave and convex models


0
2
4
6
8
10
12
14
16
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28
x(m)
y
(
m
)
ground profile
concave, r=35m
concave, r=17m
plane
convex, r=35m
convex, r=17m

Figure 4.22 Slip surfaces at different curvatures (no load)



Chapter 4:Three-dimensional Slope Failure Analysis by the Strength Reduction and Limit Equilibrium Methods

- 96 -



(a) 135 concave, FOS=1.91 (b) no curvature, FOS=1.82 (c) 145 convex, FOS=1.83
Figure 4.23 Slip surfaces for different curvatures (no load)



(a) 135 concave, FOS=1.36 (b) no curvature, FOS=1.22 (c) 145 convex, FOS=1.22
Figure 4.24 Slip surfaces for different curvatures with 200kPa loading




q=50kPa


q=50kPa

(a) convex, FOS=0.88 (b) plane, FOS=0.90
Figure 4.25 Vertical cut slope with 3m width 50kPa loading along the edge of the crest


Chapter 4:Three-dimensional Slope Failure Analysis by the Strength Reduction and Limit Equilibrium Methods

- 97 -




q=50kPa

q=50kPa

(a) convex, FOS=0.95 (b) plane, FOS=1.26
Figure 4.26 Vertical cut slope with 3m long and 3m wide 50kPa loading

Chapter 5: Soil Nailed Slope by Strength Reduction and Limit Equilibrium Methods

- 98 -

CHAPTER 5: SOIL NAILED SLOPE BY STRENGTH
REDUCTION AND LIMIT EQUILIBRIUM METHODS

5.1 Introduction
Soil nailing is a simple and economic slope stabilization technique, and is particularly
useful for the strengthening of existing slopes. There are many different design methods
for soil nailing which include the limit equilibrium methods and several working stress
design methods, and the most commonly used method for soil nailing design is still
limit equilibrium method. The Federal Highway Administration of US has also
published series of design guidelines (Elias and Christopher, 1997; Byrne et al., 1998;
Elias et al., 2001; Lazarte et al., 2003) for soil nailed and reinforced earth structures
which are based on the limit equilibrium method and are currently used by many
engineers.

The two-dimensional finite element or finite difference method has also been applied
for the analysis of nailed slope which may be more suitable for slopes reinforced with
geotextile but not soil nails. Although two-dimensional analysis is useful, the effect of
soil-nail interaction is not adequately taken into account as a soil nail has a finite
diameter with a spacing much larger than the nail diameter. Some researchers have also
employed 3D finite element method to soil nailed slope analysis, and they focus mainly
on the deformation and soil-nail interaction while the factor of safety is seldom
considered. For a soil nailed slope, there is still a lack of detailed study, in particular for
three-dimensional analysis. This chapter will try to conduct a detailed analysis for soil
nailed slope by SRM and LEM for which there are only few previous studies, and some
Chapter 5: Soil Nailed Slope by Strength Reduction and Limit Equilibrium Methods

- 99 -
interesting results are obtained which are worth further consideration and discussion for
analysis and design.

One of the important factors for a proper analysis of soil nailed slope is the
determination of the nail load. According to Hong Kongs practice (which is a special
case of the Daviss method by Shen et al. 1981a), the ultimate bond stress of a soil nail

f
is expressed as + = tan 2
f v
D c D , where D is hole diameter,
v
is effective
vertical stress on the nail, tan is frictional coefficient between the soil and the nail, c
is cohesion of the soil. Since the adhesion and friction between the soil and nail will be
less than c and tan, a factor of safety 2.0 is given to the ultimate bond stress for
design purpose. There is also another soil nail design practice in which the bond stress
is assumed to be independent of the confining/overburden stress. The laboratory and
field tests in Hong Kong have suggested that both design methods may be correct under
different cases, and the actual nail load appears to be dependent on the soil type, time,
the grouting pressure and the topography. In the following studies, the two methods of
bond load determination are considered and the SRM and LEM are compared on the
same basis for comparisons.

For LEM analysis, the effect of soil nail is considered by applying a point load provided
by the nail on the slip surface. The effective nail load is taken as the minimum of: (a)
the bond strength between grout and soil; (b) the tensile strength of the nail; (c) the
bond strength between the grout and the nail. The software Slope2000 developed by
Cheng is used for LEM analysis of soil nailed slope.

For the present SRM analysis, FLAC3D software is used. If the bending effects are not
important, the nails can be modelled by cable elements in FLAC3D because cable
Chapter 5: Soil Nailed Slope by Strength Reduction and Limit Equilibrium Methods

- 100 -
elements provide a shearing resistance (by means of the grout properties) along their
length. In FLAC3D, the system is idealized as shown in Figure 5.1, and is represented
numerically as a spring-slider system located at the nodal points along the cable axis.
The shear behavior of the grout annulus under relative shear displacement between the
cable/grout interface and the grout/rock interface as shown in Figure 5.1b is described
numerically by: the grout shear stiffness, the grout cohesive strength, the grout friction
angle, the grout exposed perimeter, and the effective confining stress
m
(Figure 5.1c).
In the above model, it is assumed that the surrounding rock is very hard and stiff and the
grout is relatively weak. In soil-nail system, the surrounding soil is relatively weak and
both the steel bar and cement grout stiffness are much higher than the soil, so in the
present study the steel bar and cement grout are taken as a whole. At the same time, a
thin layer of material with a thickness of 4.0mm surrounding the nail is used to simulate
the shearing zone. The soil nail model used in this analysis and its corresponding
meaning compared with the idealized model in FLAC3D is shown in Figure 5.2. In
SRM analysis, the failure of the soil nail system is controlled by the pullout strength
between the soil and nail or the tensile strength of the nail tendon, and this is similar to
the LEM analysis. The confining force acting around the nail is approximately
calculated by the formula
v
D 2 per unit length in LEM analysis, while it is computed
automatically in SRM analysis, and this will cause some difference, but this difference
will be small for common problems. So the results obtained by LEM and SRM
according to the above analyzing approaches will be comparable.

In this chapter, the effects of the nail head, some special results from SRM, the effects
of failure induced by external loads and the analysis of some model tests will be
discussed. These results are actually important for the proper understanding and
analysis of a nailed slope.
Chapter 5: Soil Nailed Slope by Strength Reduction and Limit Equilibrium Methods

- 101 -
5.2 Importance of nail head in analysis
Firstly, a 6m height soil nailed slope with 45 slope angle is analyzed. 8m length nails
are installed at 1.5m centers horizontally and vertically. The diameter of the steel bar is
40mm and the grout hole diameter is 100mm. The numerical model and soil properties
are shown in Figure 5.3a. The minimum factor of safety (FOS) by the Spencers method
using the simulated annealing search by Cheng (2003) is 1.2231 for pull out failure and
1.146 for face failure.

For SRM analysis, the material properties of the grouted nail are calculated considering
a combination of the stiffness of the steel bar and the cement grout, and the Youngs
Modulus of the grouted nail is determined as 45.44GPa.

In this analysis, the elastic modulus and Poisson ratio of the soil are 15MPa and 0.42
respectively, so the shear modulus of the shear zone is 5.28MPa. The shear stiffness of
the shear zone k
g
can be estimated as (Itasca 2006):
2t/D) 10ln(1
G 2
k
g
+
=

=43.1MPa
where G is shear modulus of the shear zone and equals to 5.28MPa; D is grout hole
diameter and is equal to 0.1m; t is annulus thickness and is equal to 0.004m. The
parameters used for the study are shown in Table 5.1.

The numerical model for the nailed slope without nail head is shown in Figure 5.3b with
a FOS 1.2 by the SRM (associated flow rule) which is slightly smaller than the FOS for
pull out failure by the LEM analysis but is slightly larger than that by the face failure for
LEM analysis. It has been established in the previous two chapters (also the results in
the following section) that the FOS from SRM is usually larger than that from the LEM.
Chapter 5: Soil Nailed Slope by Strength Reduction and Limit Equilibrium Methods

- 102 -
It can be concluded that the FOS by the SRM corresponds to the face failure, and this
conclusion can be further supported by the results if a nail head is modeled in the
analysis. The slip surface and the tension stress of the nail are shown in Figure 5.4. The
critical slip surfaces corresponding to the pull out failure and face failure from the LEM
as shown in Figure 5.5 are compared with the SRM result. The slip surface by the SRM
agrees well with the face failure surface from the LEM, but is greatly different from the
pullout failure surface by the LEM. Another interesting phenomenon (Figure 5.5) is that
the slip surface from the SRM for this case is virtually the same as the one when the
slope is un-reinforced (FOS=1.11). Several other examples have been tried, and it
appears that the slip surfaces by the SRM for nailed slope without nail head are usually
similar to the un-reinforced slopes. This phenomenon appears to be not noticed in the
past. More detailed study about this phenomenon will be considered in the later part of
this chapter. If a continuous elastic surface is simulated as the nail head with 0.15 m
thickness, the FOS for this model is 1.28 and the slip surface is shown in Figure 5.6, and
this failure surface is comparable to the one by LEM for pullout failure as shown in
Figure 5.7.

The results in Figures 5.4-5.7 demonstrate the effect of nail head on the failure mode
which is seldom considered in the past, and there is little previous SRM study on this
issue. For the model with no nail head, the failure mode is actually a face failure which
is similar to that in Figure 2.2a. The soil nail with no nail head is restrained by the soil
behind the failure surface while the soil failure mass is separated from the nail, so only
the bond stress within the failure mass is effective in the stabilization. For the model
with a nail head, the failure mode is a pullout failure which is similar to that in Figure
2.2b. The failure along the nail is initiated along the portion behind the failure mass, so
the bond load from the effective length behind the failure surface is effective in the
Chapter 5: Soil Nailed Slope by Strength Reduction and Limit Equilibrium Methods

- 103 -
slope stabilization. The two failure modes as found from the study of nailed slope with
and without nail head can be compared to the design of geotextile for embankment
stabilization, where the bond loads for the two sides of geotextiles have to be checked
(usually no head/anchorage to the geotextile).

To figure out the influence of the nail head with different shape and parameters, two
different types of nail head are considered. In the first model, a 0.5m width and 0.5m
height nail head with 0.15m thickness is used and its elastic modulus varies from
15MPa to 30GPa. In the second model, a continuous elastic surface is simulated as the
nail head, and its elastic modulus also varies from 15MPa to 30GPa. The results as
shown in Table 5.2 show that there is only very small difference between the two nail
head configurations. That means, there are noticeable differences for a nailed slope with
and without nail head, but the influence of the size and material parameters of the nail
head are not important (provided that the material parameters are normal and
reasonable). For example, when the nail head with 0.14m width and 0.15m height is
used (elastic modulus is 15GPa), a factor of safety of 1.25 is determined from the SRM,
and the slip surface and nail force distribution are shown in Figure 5.8.

The influence of the Poisson ratio of the soil is also investigated by varying the Poisson
ratio from 0.15 to 0.45 in the above model, and it is found this parameter has little effect
on the factor of safety of the soil nailing slope.

5.3 Soil nailed slope stability with different soil properties
In this section, a parametric study for a homogeneous soil slope with a slope height
equal to 6m and slope angle equal to 45 is conducted, and the slope geometry and soil
nail distribution is the same as the example in Figure 5.1. In the parametric study,
Chapter 5: Soil Nailed Slope by Strength Reduction and Limit Equilibrium Methods

- 104 -
different shear strength properties are used and both the LEM and SRM analyses are
carried out. The cohesion of the soil varies from 2, 5, 10 to 20 kPa while the friction
angle varies from 5, 15, 25, 35 to 45 respectively. The density, elastic modulus and
Poisson ratio of the soil are kept constant at 20kN/m
3
, 15MPa and 0.35 respectively in
all the analyses, and the results of analyses are shown in Table 5.3. For the LEM, the
Spencers method is adopted and the tolerance in the location of the critical failure
surface is 0.0001 which is good enough for the present study.

From Table 5.3, it is found that the factors of safety by the SRM and LEM are very
similar under different combinations of soil parameters, and all the FOS obtained from
the SRM (associated flow rule) are slightly larger than those obtained by LEM with a
maximum difference of 12.4 %. This phenomenon is also similar to that as obtained in
previous chapters for un-reinforced slopes. The slip surfaces for this study are shown in
Figures 5.9 to 5.11. It can be seen that when the friction angle of the soil is small, the
slip surfaces by the SRM and LEM are in good agreement. When the friction angle of
the soil becomes very large (for example, =45), the slip surfaces by the SRM are
deeper than those by the LEM.

It has been observed that there are several special cases where a combined failure
surface may occur. For example, when c=5kPa and =35 (Figure 5.12), a combined
slip surfaces given by Figure 5.12b will be detected by the use of a very fine mesh, and
such a special slip surface will not occur when the mesh size is slightly increased
(Figure 5.12a). Even though there are major differences in the two slip surfaces, the
differences in the FOS are very small, and such phenomenon does not appear for the
LEM. It appears that that the determination of a combined failure surface is controlled
by the size of the mesh in the computation. Such combined failure surface, however,
Chapter 5: Soil Nailed Slope by Strength Reduction and Limit Equilibrium Methods

- 105 -
may be not a true phenomenon as such failure surface possesses a high factor of safety
when the LEM is used. It is suspected that due to the use of extremely fine mesh, stress
concentration occurs around the grout/nail interfaces so that the combined failure
surface will come out during the nonlinear iteration analysis. Actually, it has also been
demonstrated the limitations of the nonlinear solution algorithms in evaluating the
ultimate limit states of a slope in the previous chapters. In this respect, the engineers
need to be very careful in assessing the results from the SRM, particularly for a nailed
slope using three-dimensional analysis.

5.4 Slope with different nail inclination angle
In this section, the influence of the nail inclination is discussed. The slope geometry and
the soil nail distribution is the same as the example in Figure 5.1. The nail length is 8m
and the nail inclination varied from zero to 60 degree. Both the LEM and SRM are used
and the results are compared in Table 5.4. Two different models are developed for the
SRM. In the first model, the soil nail is simulated by cable structure element which
considers only the tension effect (SRM1). In the second model, the soil nail is simulated
by pile-structure element which can consider both the tension and bending effect
(SRM2).

With the increase of nail inclination angle, the factor of safety firstly increases and then
decreases. It means that the nail inclination angle can be optimized in practice. The
maximum FOS occurs when inclined angle is between 10 to 30 degree (for LEM=10;
for SRM1=20; for SRM2=30). Although the optimized inclined angle is different for
these three methods, the variation of the FOS is very small when the inclination angle
varies within 10 to 30. There are two common views about the optimum nail
Chapter 5: Soil Nailed Slope by Strength Reduction and Limit Equilibrium Methods

- 106 -
inclination: the nail should be horizontal; the nail should be perpendicular to the slip
surface. For this example, the factor of safety is insensitive to the inclination angle
(from 10 to 30), which is basically equivalent to the inclination varies from horizontal
to nearly perpendicular to the slip surface. When the inclination is very great (say 60 or
more), the FOS will drop quickly then.

The slip surface, nail axial force distribution and bending moment distribution are
shown in Figures 5.13 to 5.15 and Table 5.4. The maximum mobilized nail axial force
increases initially and then decreases with the nail inclination, and this also explains
why the FOS increases firstly and then decreases finally. It can also be seen that some
bending effect is mobilized in the nail (Figure 5.15), especially when the nail inclination
is large (30 to 60), but the effect of the bending moment in the nail force is very small
compared with the axial nail force (Table 5.4) so that the contribution of the bending
effect to the factor of safety is also very small. This result also supports the current
method of design where the bending effect of soil nail is neglected in the analysis.

5.5 Slope with different nail length
In this section, the influence of the nail length is discussed. The slope geometry and the
soil nail distribution are the same as example 1 in previous section. The nail length
varies from 6m to 16m. Both the LEM and SRM are used and the results are compared
(shown in Table 5.5).

In general, with the increase of the nail length, the FOS will increase which is
reasonable and obvious. There are also no major differences between the results from
Chapter 5: Soil Nailed Slope by Strength Reduction and Limit Equilibrium Methods

- 107 -
the SRM and LEM. From Figure 5.16, it can also be seen with the increase of the nail
length, the slip surface gradually gets deeper.

5.6 Slope with different soil nail layout
In this section, the influence of the soil nail layout is discussed by two different cases.
The slope geometry and the soil nail distribution is also the same as example 1 in the
previous section. In case 1, the soil nails are shorter at the upper part and longer at the
bottom part (the nail lengths of the top row, middle row and bottom row are 4m, 8m and
12m respectively). In case 2, the design is reversed (the nail lengths of the top row,
middle row and bottom row are 12m, 8m and 4m respectively). Two different nail
inclination angles are also considered: zero and 20 (Figures 5.17 and 5.18). For both
nail inclination angles, the FOS for case 1 is larger than that for case 2. Since the nail
pull out strength is assumed to be defined by the overburden/confining pressure for the
results in Figures 5.17 and 5.18, such results are not surprising. To eliminate the
influence of the overburden confining pressure, two other models are developed in
which the pull out strength (5.6kN/m) is assumed to be independent of the confining
pressure and the results are shown in Figures 5.19 and 5.20. For these new models, the
FOS by case 1 is still larger than that for case 2. Case 1 appears to be a more effective
solution for enhancing the stability of slope.

Classically, case 2 is recommended for reinforced earth structure as it is better in
controlling the displacement of the slope crest. For the ultimate limit state of slope
stability, case 1 is however found to more effective as more nail forces can be
mobilized. For the reinforcement of existing slopes, the displacement control is usually
not an important issue and the failure will start to initiate at the lower part of the slope.
If longer soil nail is installed at the bottom, it will be more effective in taking up the
Chapter 5: Soil Nailed Slope by Strength Reduction and Limit Equilibrium Methods

- 108 -
loading so that it is better for the slope stability. On the other hand, if the slope is
formed by excavation, a relatively large displacement will usually induced in the slope
crest by the releasing of stresses and the failure will start from the top to bottom. If
longer nail is installed at the top, it will be more effective in controlling the
development of the displacement and the initiation of the failure. It can therefore be
concluded that the soil nail should be installed at locations where the failure start to
initiate for the maximum efficiency.

5.7 Soil nailed slope with external pressure on the top
In this section, the influence of the external pressure on SRM analysis is discussed, and
there is a major difference between the SRM and LEM in dealing with external load
which is seldom discussed in the past. The slope height, slope angle and the soil nail
distribution is also the same as example 1. A 200 kPa pressure is applied on the top of
the slope and both the SRM and LEM results for different nail length are shown in
Tables 5.6 and 5.7. Two different SRM models are developed for this study. In the first
model, the bond stress is assumed to be controlled by the confining pressure. The results
for the first model are shown in Table 5.6 and are compared with the results by LEM.
There are major differences between the results from the SRM and the LEM, and this
phenomenon can be explained that such major differences arise from the increased
confining pressure and hence the bond stress in the SRM analysis. On the other hand,
the overburden pressure on the nail from the external load is usually not considered in
the bond stress calculation (appear to be the practice for all commercial programs), so a
lower bond stress is determined from the LEM. In this respect, there is a major
difference in the soil nail design by the SRM and LEM under the action of the external
loads if the bond load is the function of the confining stresses.

Chapter 5: Soil Nailed Slope by Strength Reduction and Limit Equilibrium Methods

- 109 -
In order to eliminate the influence of the confining stresses on the bond stress, the
second model is developed in which the bond stress is assumed to be independent of the
confining pressure (bond stress=5.6 kN/m), and the results of analyses are shown in
Table 5.7. Since the basic assumption for the bond stress is the same for SRM and
LEM, the differences between the results from the LEM and SRM are smaller than the
case in Table 5.6. It can be concluded that the SRM and LEM may give greatly different
results depending on the bond load determination method, and engineers should be
aware of this difference.

5.8 Influence of nail elastic modulus
In the previous sections, the elastic modulus of the soil nail used for the analysis is
45.44 GPa, which is determined by the combination of the stiffness from the steel bar
and the cement grout (shown in Table 5.1). In the LEM, the nail stiffness is however not
required in the analysis. To investigate the influence of the nail elastic modulus on the
stability of slope, two different kinds of nail elastic moduli for two different slope
angles are considered. The first nail stiffness is 45.44 GPa which corresponds to the
grouted steel bar nail. The second nail stiffness is 4.544 GPa which approximately
corresponds to the glass fibre reinforced plastic nail which is much more flexible in
nature. Two cases are considered here, and the slope angles under consideration are 45
(same as example 1) and vertical as shown in Figure 5.21. It can be seen from Table 5.8
that the FOS for the two different nail elastic moduli are nearly the same for the first
case. On the other hand, there are noticeable effects from the soil nail stiffness which
are shown in Table 5.9 when the slope is very steep. It means that a softer soil nail is
better for the slope stability, provided that the nail pull out strength or the nail tensile
strength is not reduced. This phenomenon can be explained from the fact that more soil
movement and hence stress-redistribution is mobilized for softer nail so that the
Chapter 5: Soil Nailed Slope by Strength Reduction and Limit Equilibrium Methods

- 110 -
mobilization of the bond stress will be greater. This phenomenon can also be compared
with the extensible and inextensible design for reinforced earth design (Elias et al.,
2001). It should also be noted that a 100 kPa loading is applied at the toe of the slope to
prevent the local failure, otherwise, a slip surface passing through the soil nails will not
appear as the analysis will be controlled by the toe slope failure. The use of a strong
material or the removal of the lower solution domain in Figure 5.21 will have similar
function as the application of a 100 kPa distribution load, and the prevention of the toe
failure will only have small influence on the results of analyses.

5.9 Analysis of a vertical soil nailed wall in Seattle, Washington
Thompson and Miller (1990) described the design, construction and performance of one
of Seattles first soil nailed walls. For this vertical soil nailed wall, nails were mostly
installed at 1.8m spacing horizontally and vertically. The nail length is 10.7m, except
for the length of the top row which is 9.8m. The diameter of the drilled holes is 203mm.
Nail bars varied from 25mm to 32mm in diameter are installed at an inclination of 15
degree, though the first row on the high wall was installed at 20 degree to avoid utilities.
A typical section of the high wall is shown in Figure 5.22. In this section, the stability of
this soil nailed vertical cut will be analyzed by the SRM.

The FOS by the SRM is 1.76 while the FOS by the LEM is 1.90 (Figure 5.23). From
Figure 5.22, it can be found that the failure is located at the toe of the slope, which
means that this is actually a bearing capacity failure, so this factor of safety is not the
real FOS of the nailed wall as toe of the slope is strengthened in practice to avoid
bearing capacity failure. In order to find the FOS of the nailed wall, two additional
models are developed. In the first model, the soil shear strength is increased to a very
large value at the toe of the slope to avoid the local bearing capacity failure with a FOS
Chapter 5: Soil Nailed Slope by Strength Reduction and Limit Equilibrium Methods

- 111 -
equals 2.10 from the SRM (Figure 5.24). In the second model, a 200 kPa pressure is
applied at the toe of the slope and the FOS is 2.29 from the SRM (Figure 5.25). In these
two additional models, two failure modes are detected which are much clearer in the
second model (Figure 5.25). In the first failure mode, the upper part of the slip surface is
nearly vertical which is basically the inextensible strip failure mode as stated in FHWA
(Elias et al., 2001). In the second failure mode, the slip surface is nearly the same as the
one by the LEM and is approaching that for the extensible strip failure mode. Since the
inextensible strip failure mode is not obtained from the LEM analysis, the SRM have
some advantage over the LEM for this case.

5.10 Distribution of the nail tension force and critical slip surface
The line of maximum tension within the nail is often considered as the failure surface
which divides the soil mass into two regions: (1) active zone, which is near to the
ground surface and in this region the shear stress mobilized on the nail surface is toward
outside and tend to pull the nail out of the soil; (2) resistant zone, which is behind the
slip surface and in this region the shear stress generated on the nail is toward inside and
tend to prevent the nail from being pulled out. This concept is shown in Figure 5.26
(after Byrne et al., 1998) and many reinforced earth structures are designed in this way.
It should be realized that the maximum tensile force location is not necessarily the
traditional critical failure surface at the limit state, but is the consequence of the
interaction between soil and nail (Byrne et al., 1998). Classically, many researchers
assume that the maximum tensile force line coincides with the potential sliding surface,
and such view is supported by some model tests (J uran et al., 1984) and full-scale tests
(Clouterre, 1993). In this section, this issue will be discussed by conducting strength
reduction analysis of several soil nailed slopes.

Chapter 5: Soil Nailed Slope by Strength Reduction and Limit Equilibrium Methods

- 112 -
Firstly, the influence of the failure modes is investigated. The soil nailed slope failure
can be broadly classified into external failure modes and internal failure modes. For
external failure modes, the slip surface does not intersect with the nails, so the
maximum tensile force line will not coincide with the potential sliding surface. Three
different external failure modes are shown in Figure 5.27. In Figure 5.27a, if the soil
nail is very short, it will be a global failure where the nails are totally within the failure
mass. In Figure 5.27b, the soil nail is long and the soil mass/nail becomes an integral
body and fails by a sliding mode. In Figure 5.27c which is one of Seattles first soil
nailed walls discussed in the previous section, the reinforced soil nailed wall is more
stable than the toe of the slope so that it is a local bearing capacity failure. For internal
failure modes, they are usually classified into three different types face failure, pullout
failure and nail tensile failure which are shown in Figure 2.2 (after Byrne et al., 1998).
These three different internal failure modes are modeled for the slope as shown in
Figure 5.1, and the results are shown in Figure 5.28. In Figure 5.28a which is a face
failure, no nail head is used and the soil nail is restrained by the soil mass behind the
failure surface so that only the friction within the failure mass is effective in the
stabilization. In this case, the line of the maximum tension force is located behind the
slip surface. In Figure 5.28b, the line of the maximum tension force is near to the nail
head and is in front of the slip surface. In Figure 5.28c where the tensile strength of the
nail is only 10kN, the nail will reaches its tensile strength at ultimate limit state and the
line of the maximum tension force will virtually coincides with the slip surface.

Secondly, the influence of the state of the slope (service state and limit state) is
investigated. Slip surface should be referred to the condition of the failure mass at the
limit state. For the maximum tension force line, it can be referenced at the service state
or the limit state which should be stated clearly. To consider this, a model with 12 m
Chapter 5: Soil Nailed Slope by Strength Reduction and Limit Equilibrium Methods

- 113 -
length soil nail with nail head is developed to compare these two states and the slope
geometry and the soil nail distribution is the same as example 1. The results of analysis
are shown in Figure 5.29. In Figure 5.29a (service state), the slip surface at the limit
state is shown by the dash line for comparison. It can be seen that the line of the
maximum tension force at service state is behind the slip surface. The force at the nail
head connection at service state is very small. In the limit state (Figure 5.29b), the line
of the maximum tension force is in front of the slip surface, and a relatively large force
at the nail head connection is mobilized at the limit state. With the reduction of the soil
shear strength, the soil nailed slope gradually transforms from service state to limit
state, and during this process the nail force is gradually mobilized while the line of
maximum nail force will gradually move towards the nail head. In this example which
is an internal pullout failure, the slip surface appears to be located between the lines of
maximum nail force at the service state and the limit state. Another example is shown in
Figure 5.30 which is the Seattles first soil nailed walls discussed in the previous
section. For this example, during the transition from the service state to the limit state,
the nail force is gradually mobilized and the line of the maximum nail force also
gradually moved towards the nail head. Compared with Figures 5.24 and 5.25, it is clear
that the line corresponding to the maximum tension is far away from the slip surface,
but is very similar to the inextensible strip design method by FHWA (Elias et al., 2001).
In the above analysis, the limit state is achieved by reducing the shear strength which is
easy to be conducted in numerical simulation but not for model tests or full-scale tests.
The limit state can also be achieved by applying pressure on top of a slope, and an
example is shown in Figure 5.31 where the results at the limit state are achieved by
different ways. In Figure 5.31a, the limit state is achieved by reducing soil shear
strength and the nail force at the bottom row is slightly more mobilized than the upper
two rows. In Figure 5.31b, the limit state is achieved by applying a pressure near to the
Chapter 5: Soil Nailed Slope by Strength Reduction and Limit Equilibrium Methods

- 114 -
crest of the slope with an internal failure, and the nail force at the top row is more
mobilized. In Figure 5.31c, the limit state which is an external failure is achieved by
applying a pressure on top of the slope near to end of the soil nails, and the nail force at
the bottom row is very large while the nail forces at the upper two rows are much lower.

Some major test results in literature are summarized in Figure 5.32 and Figure 5.33. In
Figure 5.32a (Gassler, 1993) and Figure 5.32b (Stocker et al., 1979), the limit state is
achieved by applying a pressure on top of the slope. In Figure 5.32c (Clouterre, 1993),
the limit state is achieved by gradually saturating the soil and this can be approximately
viewed as equivalent to reducing the soil strength. At the limit state, the slip surface can
be observed and both the slip surface and the nail tension force distribution are given in
Figure 5.32. In Figures 5.32a and 5.32b, both the slip surfaces only partly intersect the
soil nails. Actually, the limit state is caused by the applied load on top of the slope. The
location of the slip surface is controlled mainly by the location of the loading as the slip
surfaces starts from the edge of the loading (similar to those in Figure 5.31b and 5.31c).
Since the location of the slip surfaces are controlled mainly by the location of the
loading in Figures 5.32a and 5.32b and the slip surface only partly intersect the soil
nails, the line of the maximum tension force will not coincides with the slip surface, and
it can be seen that the maximum tension force at the lower rows of nails is very close to
the nail head. In Figure 5.32c where the ultimate state is controlled by the shear strength
reduction, it is an internal tension failure and the line of the maximum tension force
coincides well with the slip surface. In this test, the structure had been designed with a
sufficiently low safety factor for failure by the breakage of the nails (F=1.1). When the
soil is gradually saturated, the system failed by the breakage of the nail which is similar
to the SRM results in the previous section.

Chapter 5: Soil Nailed Slope by Strength Reduction and Limit Equilibrium Methods

- 115 -
There are also four different soil nailed test results at service state shown in Figure 5.33.
In service state, the slip surface does not exist so that only the tension force distribution
is presented in these figures (in Figure 5.33b, two predicted slip surfaces are also
shown). In Figure 5.33a (Shen et al., 1981c), the measured maximum tension force is
near to the nail head. In Figure 5.33b, 5.33c and 5.33d, the maximum tension force at
the lower part of the slope is also very close to the nail head while and the maximum
tension force at the upper part is slightly further away from the nail head.

Based on the above analysis, it can be concluded that the line of the maximum tension
does not correspond to the traditional critical slip surface in general, but is the
representation of the soil structure interactions in the soil nailing system. Firstly, the
tension stress distribution is influenced by the state of the slope and it is obviously
different between the service state and the limit state. When the soil nailed slope
gradually transforms from the service state to the limit state by reducing the shear
strength, if face failure is prevented by the use of nail head, the maximum tension line
will move towards the slope surface (such as Figure 5.29 and Figure 5.30) as the nail
force is gradually mobilized and the resistant zone gets larger to maintain the slope in
stable condition. Secondly, the tension stress distribution is influenced by the slope
failure modes at limit state. For slope with internal failure modes, if it is face failure, the
line of maximum tension usually is located behind the slip surface (Figure 5.28a). For a
pullout failure, the line of the maximum tension is usually located in front of the slip
surface (Figure 5.28b, Figure 5.29b). For a nail tensile failure, the line of the maximum
tension will coincides well with the slip surface which is demonstrated by both the
numerical simulation (Figure 5.28c) and the full scale test (Figure 5.32c). For slope with
external failure modes, the line of the maximum tension force will not coincide with the
Chapter 5: Soil Nailed Slope by Strength Reduction and Limit Equilibrium Methods

- 116 -
slip surface. Besides, both the nail tension force distribution and the slip surface are
greatly controlled by the location of the external loading (Figure 5.31).

5.11 Slip surface for face failure
As presented in the previous section, for the soil nailed slope shown in Figure 5.1, if the
nail head is not modeled, face failure will occurs and the slip surface by the SRM
happened to be virtually the same as the one when the slope is not reinforced (though
the FOS are not the same). Since such phenomenon is found for many cases (not shown
in this chapter), the validity of this observation under different cases will be
investigated. In this section, three different models are considered. In the first model,
the horizontal and vertical interval of the nail is 1.5m (three nails are included in the
model) which is the same as example 1 in previous sections. In the second model, the
nail interval is 1.0m (five nails in the model) while the nail interval is 0.75m in the third
model (eight nails in the model). At the same time, two different soil properties are
considered: a soil with c=9 kPa and =18 (FOS=1.11 with no nail); the second case
where c=20 kPa and =5 (FOS=1.2 with no nail). The results are shown in Figure
5.34 and 5.35 and the slip surface with no nail is shown in dashed line. It can be seen
that when the nail interval is large, the slip surface is nearly the same with the one with
no nail. When the nail interval gets smaller, the difference becomes obvious. For the
first case where is greater, the slip surface gets shallower with the decrease of the nail
interval while the slip surface gets deeper with the decrease of the nail interval for the
second case. It is however true that for normal nail spacing, the location of the slip
surface for face failure is relatively insensitive to the presence of soil nail.


Chapter 5: Soil Nailed Slope by Strength Reduction and Limit Equilibrium Methods

- 117 -
5.12 Influence of nail spacing on the failure modes
For the SRM analysis in previous sections, though the 3D analyzing is conducted for
better consideration of the soil-nail interaction, the slip surfaces are still basically two
dimensional (the slip surface is nearly the same at different section). This phenomenon
is due to the relatively small nail spacing used in the analysis. If the spacing of the nail
is large enough, a clear three dimensional failure surface may occur. A slope model is
developed in this section to illustrate the influence of the nail interval on the failure
mode. The slope geometry and soil property is the same as the example shown in Figure
5.1, but the horizontal and vertical intervals of the nail are 10m and 1m respectively
(five nails in the model), and the nail length is 14m. When the soil-nail interface
strength is set to be half of the soil shear strength, the slip surface is still nearly two
dimensional (Figure 5.36). When the soil-nail interface strength is set to be the same as
soil shear strength, a clear three dimensional slip surface will be mobilized (Figure
5.37). This means that the failure mode is influenced by both the nail interval and the
load provided by the nail. Since the basic idea of the soil nailing technique is installing
close spaced inclusions into the soil to increase the stability, a clear three-dimensional
slip surface would not be easily mobilized simply due to the presence of nail. On the
other hand, if the slope is reinforced by anchor technique which can provides large
supporting load by each individual reinforcing element with a relatively large spacing, a
clear three-dimensional slip surface may be easier to be mobilized.

5.13 Discussion and Conclusion
In this chapter, it is established that the Poisson ratio and the stiffness/arrangement of
the nail head have little influence on the factor of safety and the failure mechanism of a
slope. The presence of a nail head is however important and should be properly
modeled. There have been slope failures in Hong Kong where the reason can be
Chapter 5: Soil Nailed Slope by Strength Reduction and Limit Equilibrium Methods

- 118 -
attributed to the use of very small nail head (and hence face failure) so that this case
should be checked in the analysis and design. The option of face failure is however
absent in many commercial slope stability programs, and engineers should provide an
adequate nail head in general for maximum efficiency. As long as the nail head is not
too small, a pull out failure will be the failure mechanism which is investigated in the
present study.

From the present study, it is found that the FOS from the SRM and the LEM are similar
under most cases, and nearly all the FOS from the SRM are slightly greater than those
from the LEM. Although the slip surfaces by the SRM and the LEM usually agree well,
sometimes the slip surface from the SRM is not regular which is one of the limitations
of the SRM. From the analysis of a soil nailed wall in Seattle, it is also found that
several failure modes can easily be detected by the SRM which are absent in the
corresponding LEM analyses. In this respect, the SRM can be an alternative or
complementary method to the LEM analysis. On the other hand, some special combined
failure surfaces similar to that in Figure 5.12b are found when a very fine mesh is used
in the analysis. In this respect, the SRM also possesses numerical problems not found in
the LEM.

It is found that the nail elastic modulus have little influence on the SRM analyses
except for very steep slope. The FOS from the SRM also appears to increase with the
decreasing of nail elastic modulus which can be attributed to the greater soil movement
and mobilization of nail bond load. This phenomenon can also account for the
differences in the design of reinforced earth wall using the extensible and inextensible
strip methods (Elias et al., 2001) which is based on field observations but are also
obtained from numerical modeling in the present study.
Chapter 5: Soil Nailed Slope by Strength Reduction and Limit Equilibrium Methods

- 119 -
When there is an external pressure on top of a slope and the bond stress is assumed to
be dependent on the overburden/confining stress, there are great differences between the
SRM and LEM. In all the commercial slope stability programs based on the LEM, the
increase in the overburden stress from the external loads on the nails are not included as
there is not a reliable method in distributing the external loads under ultimate limit state.
The simplified 1 in 2 rule of thumb or other empirical methods are all based on the
service state instead of the ultimate limit state and are not included in normal LEM
analysis. On the other hand, SRM can consider the external load easily and
automatically which is an advantage not present in LEM. Engineers should also be
aware of this difference in the bond load determination in LEM and SRM, and such
difference has seldom been considered up to present.

In soil nailing design, the nail inclination angle and the layout can be optimized by the
SRM or LEM analyses. Usually, with the increase of the nail inclination angle, the FOS
increases slightly initially which will decreases if the inclination is large. An optimum
soil nail design can be determined easily using whether the SRM or the LEM. It is also
found that the mobilized bending moment in soil nail is small even when the nail
inclined angle is large. The use of the bond stress for soil nail design which is a
commonly adopted approach appears to be a reasonable design method.

An optimum design of longer nail at top and shorter nail at bottom has been
recommended by some researchers and engineers for the control of the movement for
soil nails installed by a top-down construction. For the stabilization of existing slopes,
from the safety factor point of view, the SRM analysis shows that the reverse
arrangement will be a more economic solution for the ultimate limit state. The final
conclusion to the optimum soil nail layout from the present study is to place a longer
Chapter 5: Soil Nailed Slope by Strength Reduction and Limit Equilibrium Methods

- 120 -
nail at locations where failure will start to initiate, and the choice of longer or shorter
nail at top or bottom will depend on the initiation of the possible failure mechanism.

The tension force distribution along soil nail is found to be controlled by the state of the
slope (service state, limit state) and the failure modes (external failure, internal failure).
In general, the line of maximum tension does not correspond to the conventional critical
slip surface which is different from the common believe, but is the consequence of the
soil nail interaction. For slope with external failure modes, the line of maximum tension
force will never coincide with the slip surface. For slope with internal face failure
modes, the line of maximum tension is usually found to locate behind the slip surface.
For a pullout failure, the line of maximum tension is usually in front of the slip surface
while for a tensile failure, the line of maximum tension will coincides with the slip
surface as the common belief.

The limit state can be achieved by reducing the soil shear strength or by applying
external loading. It is demonstrated that there are differences in the nail force
distribution and slip surface between these two modes. When the soil nailed slope
gradually transforms from service state to limit state by reducing the shear strength, if
the nail head is strong enough with no face failure, the maximum tension force line will
move towards the slope surface.

The failure mode of soil nailed slope is controlled by both the nail spacing and the nail
load. In practice, the nail is usually closely spaced, thus a clear three-dimensional slip
surface would usually not easy to be mobilized if the slope geometry and boundary
conditions have no obvious 3D effects. Most of the failure modes examined in this
chapter are basically two dimensional slip surfaces, as the nail is densely populated and
Chapter 5: Soil Nailed Slope by Strength Reduction and Limit Equilibrium Methods

- 121 -
the slope has no distinct irregular geometry or loading. In the next chapter, some soil
nailing slopes with distinct 3D effect in geometry and boundary conditions will be
investigated.























Chapter 5: Soil Nailed Slope by Strength Reduction and Limit Equilibrium Methods

- 122 -

Table 5.1 Parameters of grout-soil-nail system
Youngs modulus, E [GPa]
45.44
grout cohesive strength (force) per unit length, c
g
[kN/m] 1.4135
grout friction angle,
g
[degree] 9.23
grout stiffness per unit length, k
g
[MPa] 43.1
grout exposed perimeter, p
g
[m] 0.339
cross-sectional area, A [m
2
] 0.00785
compressive yield strength (force), F
c
[MN] 0.238
tensile yield strength (force), F
t
[MN] 0.238

Table 5.2 Factor of safety for nail head with different elastic modulus
Elastic modulus 30GPa 15GPa 15MPa
FOS (0.5m*0.5m nail head) 1.28 1.27 1.28
FOS (continuous nail head) 1.26 1.27 1.27

Table 5.3 Factors of safety by LEM and SRM
case
c
(kPa)

()
factor of
safety
(LEM)
factor of
safety
(SRM)
FOS difference
between LEM
and SRM (%)
1 2 5 0.27 0.28 3.70
2 2 15 0.59 0.63 6.78
3 2 25 0.99 1.09 10.10
4 2 35 1.55 1.72 10.97
5 2 45 2.34 2.63 12.39
6 5 5 0.44 0.46 4.55
7 5 15 0.80 0.86 7.50
8 5 25 1.23 1.33 8.13
9 5 35 1.80 1.96 8.89
10 5 45 2.59 2.9 11.97
11 10 5 0.70 0.74 5.71
12 10 15 1.14 1.19 4.39
13 10 25 1.63 1.71 4.91
14 10 35 2.17 2.37 9.22
15 10 45 2.97 3.3 11.11
16 20 5 1.15 1.26 9.57
17 20 15 1.67 1.79 7.19
18 20 25 2.20 2.34 6.36
19 20 35 2.82 3.04 7.80
20 20 45 3.67 3.95 7.63

Chapter 5: Soil Nailed Slope by Strength Reduction and Limit Equilibrium Methods

- 123 -


Table 5.4 Results for different nail inclination
Inclination angle () 0 10 20 30 40 60
FOS by LEM 1.22 1.26 1.24 1.24 1.22 1.13
Maximum force of top nail
by LEM (kN)
3.71 6.54 11.26 14.17 15.63 19.00
Maximum force of middle
nail by LEM (kN)
5.9 9.21 13.85 17.06 18.80 21.54
Maximum force of bottom
nail by LEM (kN)
8.74 12.39 17.28 20.47 21.72 23.10
FOS by SRM1 1.28 1.30 1.33 1.33 1.31 1.17
Maximum force of top nail
by SRM1 (kN)
13.85 13.22 14.78 15.27 15.94 3.79
Maximum force of middle
nail by SRM1 (kN)
17.22 21.55 26.92 30.25 33.00 23.33
Maximum force of bottom
nail by SRM1 (kN)
28.15 31.41 35.80 39.57 41.69 24.65
FOS by SRM2 1.28 1.30 1.33 1.36 1.36 1.20
Maximum force of top nail
by SRM2 (kN)
18.52 19.56 21.54 23.79 26.22 4.99
Maximum force of middle
nail by SRM2 (kN)
21.35 26.49 28.77 33.46 36.15 22.13
Maximum force of bottom
nail by SRM2 (kN)
30.32 34.98 38.84 41.59 44.76 23.77
Maximum moment (kNm) 1.193 1.215 1.079 1.744 1.820 1.819


Table 5.5 Factor of safety for different soil nail length
Nail length (m) 0 (no nail) 6 8 10 12 14 16
FOS by SRM 1.11 1.17 1.28 1.36 1.41 1.48 1.54
FOS by LEM 1.05 1.12 1.22 1.27 1.32 1.35 1.40


Table 5.6 Factor of safety with 200 kPa top pressure (bond load controlled by
overburden stress)
Nail length (m) 8 12 16
FOS by SRM 0.66 0.76 0.84
FOS by LEM 0.56 0.59 0.61


Chapter 5: Soil Nailed Slope by Strength Reduction and Limit Equilibrium Methods

- 124 -


Table 5.7 Factor of safety with 200 kPa top pressure (constant pull out resistance)
Nail length (m) 8 12 16
FOS by SRM 0.60 0.67 0.73
FOS by LEM 0.58 0.61 0.65


Table 5.8 Factor of safety with different nail elastic modulus (slope angle 45 degree)
Nail length (m) 8 12 16
FOS by SRM (E=45.44 GPa) 1.28 1.41 1.54
FOS by SRM (E=4.544 GPa) 1.29 1.44 1.57


Table 5.9 Factor of safety with different nail elastic modulus (vertical cut slope)
Nail length (m) 6 8 12
FOS by SRM (E=45.44 GPa) 1.32 1.51 2.12
FOS by SRM (E=4.544 GPa) 1.57 1.79 2.37













Chapter 5: Soil Nailed Slope by Strength Reduction and Limit Equilibrium Methods

- 125 -



Figure 5.1 Idealization of grout-cable system (from Itasca, 2006)


Grouted nail (cable)
Shear zone (grout)
Soil

Figure 5.2 Idealization of soil nail system


Chapter 5: Soil Nailed Slope by Strength Reduction and Limit Equilibrium Methods

- 126 -

0,0
0,5
6,5
28,0
12,11
28,11
c=9kPa
=18
Density=19kN/m
3

X
Z
7.5,6.5
9.0,8.0
10.5,9.5
Soil nail

(a) cross section plot of the soil nailed slope model
X
Z
Y

(b) three-dimensional mesh for SRM analysis
Figure 5.3 Plot of the soil nailed slope model



Figure 5.4 Slip surface and the tension stress of soil nail without nail head (FOS=1.20)



Chapter 5: Soil Nailed Slope by Strength Reduction and Limit Equilibrium Methods

- 127 -

LEM with no soil nail
LEM with soil nail (pull out failure)
LEM with soil nail (face failure)

Figure 5.5 Slip surface for soil nailed slope without nail head by SRM and compared
with LEM results



Figure 5.6 Slip surface and the tension stress of soil nail for model with nail head
(FOS=1.28)



LEM with no soil nail
LEM with soil nail (pull out failure)

Figure 5.7 Slip surface for soil nailed slope by SRM for the model with nail head

Chapter 5: Soil Nailed Slope by Strength Reduction and Limit Equilibrium Methods

- 128 -


Figure 5.8 Slip surface of the slope with 0.14m width and 0.15m height nail head

0
2
4
6
8
10
12
0 2 4 6 8 10 12 14 16 18 20 22 24
x(m)
y
(
m
)
ground profile and nail
limit equilibrium
SRM
0
2
4
6
8
10
12
0 2 4 6 8 10 12 14 16 18 20 22 24
x(m)
y
(
m
)
ground profile and nail
limit equilibrium
SRM

(a) c=2kPa, =5 (b) c=2kPa, =15
0
2
4
6
8
10
12
0 2 4 6 8 10 12 14 16 18 20 22 24
x(m)
y
(
m
)
ground profile and nail
limit equilibrium
SRM
0
2
4
6
8
10
12
0 2 4 6 8 10 12 14 16 18 20 22 24
x(m)
y
(
m
)
ground profile and nail
limit equilibrium
SRM

(c) c=2kPa, =25 (d) c=2kPa, =45
Figure 5.9 Slip surface comparison with increasing friction angle (c=2kPa)


Chapter 5: Soil Nailed Slope by Strength Reduction and Limit Equilibrium Methods

- 129 -

0
2
4
6
8
10
12
0 2 4 6 8 10 12 14 16 18 20 22 24
x(m)
y
(
m
)
ground profile and nail
limit equilibrium
SRM

0
2
4
6
8
10
12
0 2 4 6 8 10 12 14 16 18 20 22 24
x(m)
y
(
m
)
ground profile and nail
limit equilibrium
SRM

(a) c=5kPa, =5 (b) c=5kPa, =15
0
2
4
6
8
10
12
0 2 4 6 8 10 12 14 16 18 20 22 24
x(m)
y
(
m
)
ground profile and nail
limit equilibrium
SRM
0
2
4
6
8
10
12
0 2 4 6 8 10 12 14 16 18 20 22 24
x(m)
y
(
m
)
ground profile and nail
limit equilibrium
SRM

(c) c=5kPa, =25 (d) c=5kPa, =45
Figure 5.10 Slip surface comparison with increasing friction angle (c=5kPa)

0
2
4
6
8
10
12
0 2 4 6 8 10 12 14 16 18 20 22 24
x(m)
y
(
m
)
ground profile and nail
limit equilibrium
SRM

0
2
4
6
8
10
12
0 2 4 6 8 10 12 14 16 18 20 22 24
x(m)
y
(
m
)
ground profile and nail
limit equilibrium
SRM

(a) c=20kPa, =5 (b) c=20kPa, =15
0
2
4
6
8
10
12
0 2 4 6 8 10 12 14 16 18 20 22 24
x(m)
y
(
m
)
ground profile and nail
limit equilibrium
SRM

0
2
4
6
8
10
12
0 2 4 6 8 10 12 14 16 18 20 22 24
x(m)
y
(
m
)
ground profile and nail
limit equilibrium
SRM

(c) c=20kPa, =25 (d) c=20kPa, =45
Figure 5.11 Slip surface comparison with increasing friction angle (c=20kPa)


Chapter 5: Soil Nailed Slope by Strength Reduction and Limit Equilibrium Methods

- 130 -





0
2
4
6
8
10
12
0 2 4 6 8 10 12 14 16 18 20 22 24
x(m)
y
(
m
)
ground profile and nail
limit equilibrium
SRM

(a) FOS=1.96 for slightly increased element size (b) FOS=1.94 for very fine mesh
Figure 5.12 Slip surface obtained by different element size (c=5kPa, =35)




(a) nail inclination angle =0 (b) nail inclination angle =10

(c) nail inclination angle =30 (d) nail inclination angle =60
Figure 5.13 Slip surface and nail axial force distribution for different nail inclined
angle (nail simulated by cable element)






Chapter 5: Soil Nailed Slope by Strength Reduction and Limit Equilibrium Methods

- 131 -


(a) nail inclination angle =0 (b) nail inclination angle =10

(c) nail inclination angle =30 (d) nail inclination angle =60
Figure 5.14 Slip surface and nail axial force distribution for different nail inclination
(bending effect considered)




(a) nail inclination angle =0 (b) nail inclination angle =10

(c) nail inclination angle =30 (d) nail inclination angle =60
Figure 5.15 Slip surface and nail bending moment distribution for different nail
inclination (bending effect considered)




Chapter 5: Soil Nailed Slope by Strength Reduction and Limit Equilibrium Methods

- 132 -



(a) L=6m (b) L=10m

(c) L=14m (d) L=16m
0
2
4
6
8
10
12
0 2 4 6 8 10 12 14 16 18 20 22 24
x(m)
y
(
m
)
ground profile and nail
6m
8m
10m
12m
14m
16m

(e) Slip surface location comparison for different nail length
Figure 5.16 Slip surface and the nail tension force for different nail length



Maximumnail force=42.2kN Maximumnail force=22.1kN

(a) Short nail at top, FOS=1.31 (b) short nail at bottom, FOS=1.22
Figure 5.17 Slip surface and the tension stress in different layout with zero inclined
angle



Chapter 5: Soil Nailed Slope by Strength Reduction and Limit Equilibrium Methods

- 133 -



Maximumnail force=60.9kN

Maximumnail force=28.3kN

(a) short nail at top, FOS=1.40 (b) short nail at bottom, FOS=1.26
Figure 5.18 Slip surface and the nail load for different layout with 20 nail inclination



Maximumnail force=46.4kN

Maximumnail force=30.2kN

(a) short nail at top, FOS=1.34 (b) short nail at bottom, FOS=1.22
Figure 5.19 Results for different layout with zero inclined angle (constant nail pullout
strength)



Maximumnail force=46.4kN
Maximumnail force=30.7kN

(a) short nail at top, FOS=1.35 (b) short nail at bottom, FOS=1.26
Figure 5.20 Results for different layout with 20 nail inclination (constant nail pullout
strength)




Chapter 5: Soil Nailed Slope by Strength Reduction and Limit Equilibrium Methods

- 134 -


X
Y
0,0
0,5
6,5
26,0
6,10.5
26,13
Cohesion=18kPa
Friction angle=25
Density=19kN/m
3

6,6
6,7.5
6,9
Soil nail
6,12
6,13
100 kpa

Figure 5.21 Vertical cut soil nailed slope model



Local failure

Figure 5.22 Slip surface and nail tension stress of the vertical soil nailed wall with
FOS=1.76



Figure 5.23 Slip surface of the vertical soil nailed wall by LEM with FOS=1.90

Chapter 5: Soil Nailed Slope by Strength Reduction and Limit Equilibrium Methods

- 135 -

Slip surface
by LEM
Soil with very high
shear strength

Figure 5.24 Slip surface of the vertical soil nailed wall with high shear strength at left
part by SRM (FOS=2.10)

Failuremode1
Failuremode2
Slip surface
by LEM
200 kpa

Figure 5.25 Slip surface of an excavated vertical cut with applied pressure 200kPa at
left corner by SRM (FOS=2.29)

Resistant zone
Activezone
Facing
Distribution of tension
force along nail
Line of maximumtension
force

Figure 5.26 Load transfer mechanism in soil nails (after Byrne et al., 1998)

Chapter 5: Soil Nailed Slope by Strength Reduction and Limit Equilibrium Methods

- 136 -


Local failure

(a) global failure (b) sliding failure (c) bearing capacity failure
Figure 5.27 Axial tensile force distribution of soil nail in different external failure
modes


Locus of maximumtension

(a) no nail head, FOS=1.20, face failure (b) strong nail head, FOS=1.28, pullout failure

(c) nail head is strong, nail tensile strength=10kN, FOS=1.22 with nail tensile failure
Figure 5.28 Slip surface and nail tension stress distribution in different internal failure
modes


Slip surface at limit state
Locus of maximumtension forces
Locus of maximumtension forces

(a) service state (b) limit state, FOS=1.41
Figure 5.29 Axial tensile force distribution of soil nail in different state


Chapter 5: Soil Nailed Slope by Strength Reduction and Limit Equilibrium Methods

- 137 -



Locus of maximumtension forces

Locus of maximumtension forces

(a) service state (b) limit state
Figure 5.30 Axial tensile stress distribution in different state for a vertical cut slope




50kPa

(a) limit state by reducing strength, f=1.31 (b) limit state by applying pressure near crest
120kPa

(c) limit state by applying 120kpa pressure at slope top near the nail end
Figure 5.31 Axial tensile stress distribution in different limit state




Chapter 5: Soil Nailed Slope by Strength Reduction and Limit Equilibrium Methods

- 138 -






50
40
30
20
10
0
(kN)
40
30
20
10
0
(kN)
50
40
30
20
10
0
(kN)
50
40
30
20
10
0
(kN)
40
30
20
10
0
(kN)
1
2
3
4
5
150 100
60
0
150
100 60
0
150
100
60
0
150
100
60
0
110
100
60
0
150
110
110
Slip surface

Failure surface
Line of maximum
tension force
Tensile force
20kN
10kN

(a) Gassler (1993) (b) Stocker et al. (1979)
Observed crack
Tensile stresses
distribution
Slip surface
20
10
20
10
20
10
20
10
20
10
T(kN)

(c) Clouterre (1993)
Figure 5.32 Nail force distribution in limit state





Chapter 5: Soil Nailed Slope by Strength Reduction and Limit Equilibrium Methods

- 139 -






3rd level anchor rod
2nd level anchor rod
4th level anchor rod
H
wall
=30ft
H
wall
=24ft
H
wall
=48ft
H
wall
=30ft
H
wall
=24ft
H
wall
=30ft
Distance behind wall facing (ft.)
0 5
10 15 20
0
1
2
3
4
0
1
2
3
4
0
1
2
3
4
5
A
x
i
a
l

f
o
r
c
e

(
k
i
p
s
)


KIPS
Reinforced Earth
Method
kN
LOAD
100
20
0
0
Davis Method
Predicted Locus of Maximum Tensile Force

(a) Shen et al. (1981c) (b) Thompson and Miller (1990)
Scaleof stresses
=40 kPa
Tensilestresses distribution
Attempt for delimitation between
activeand passivezone

N1
N2
N3
N4
N1
N2
N3
N4
Nail length (m)
N
a
i
l

f
o
r
c
e

(
k
N
)

60
40
20
0
1 2 3 4 5 6 7 8

(c) Cartier and Gigan (1983) (d) Stocker and Riedinger (1990)
Figure 5.33 Nail force distribution in service state






Chapter 5: Soil Nailed Slope by Strength Reduction and Limit Equilibrium Methods

- 140 -




(a) nail interval 1.5m, FOS=1.20 (b) nail interval 1.0m, FOS=1.35

(c) nail interval 0.75m, FOS=1.65
Figure 5.34 Slip surface of soil 1 with face failure in different nail interval




(a) nail interval 1.5m, FOS=1.32 (b) nail interval 1.0m, FOS=1.49

(c) nail interval 0.75m, FOS=1.75
Figure 5.35 Slip surface of soil 2 with face failure in different nail interval



Chapter 5: Soil Nailed Slope by Strength Reduction and Limit Equilibrium Methods

- 141 -



(a) slip surface (b) nail force distribution
Figure 5.36 Slip surface and nail force for slope with 10m horizontal nail interval and
soil-nail interface strength is half of the soil strength (FOS=1.25)



(a) slip surface (b) nail force distribution
Figure 5.37 Slip surface and nail force for slope with 10m horizontal nail interval and
soil-nail interface strength is the same as the soil strength (FOS=1.30)
Chapter 6: Effects of Curvature and Local Loading on 3D Soil Nailed Slope

- 142 -

CHAPTER 6: EFFECTS OF CURVATURE AND LOCAL
LOADING ON 3D SOIL NAILED SLOPE

6.1 Introduction
In traditional engineering design, soil nails are usually designed using 2D analysis
which is reasonable when the soil nails are densely installed and the slope has no
obvious 3D effect in geometry and boundary conditions. In most cases, 3D slope failure
will be the prevailing failure mode for real problems, and chapter 4 has carried out
detailed study on the 3D failure of unreinforced slopes. In this chapter, 3D analysis will
be conducted for some soil nailed slopes with strong 3-D effect (locally loaded slope,
curved slope and intersected slope), and the factor of safety of the slope will be obtained
by the strength reduction method using software FLAC3D.

6.2 Numerical simulation for locally loaded soil nailing slope
In this section, a 6m height soil nailing slope with 45 slope angle under rectangular
shape vertical loading as shown in Figure 6.1 is analyzed. The loading width B is 2m
while the edge of the loading is 1m away from the crest of the slope. Nails are installed
at 1.5m centers horizontally and vertically. The nail length is 8m and the inclination
angle is zero. The slope geometry and the soil properties are shown in Figure 6.1, while
the numerical model is 18m in length, so there are 36 nails included in this model (3
row and 12 nails in each row). When the loading q is 200kPa, the results of analysis are
shown in Figure 6.2, where the ratio of the loading length L to loading width B varies
from 1 to 4.

Chapter 6: Effects of Curvature and Local Loading on 3D Soil Nailed Slope

- 143 -
When the L/B ratio is 1.0, the failure is a local failure controlled by the top surcharge
and does not pass through the toe of the slope. With the increase of the loading length
(L/B increases from 1 to 4), the factor of safety becomes smaller and the slip surface
becomes larger and extends towards the toe of the slope. The maximum nail tension
force distribution at limit state under different local loading is shown in Figures 6.3 to
6.5, and the maximum nail force distribution for the slope with no local loading is
shown in Figure 6.6. If no loading is applied on the nailing slope, the nail force at the
bottom row is much more mobilized than the force at top and middle rows at the limit
state (Figure 6.6). The tension force for the middle row is also slightly more mobilized
than the top row, but the difference in the nail force between the middle row and top
row is not large. The nail force of the bottom row is much larger than the middle and the
top row. At the limit state, the movement of the slope is larger at the lower part and is
smaller at upper part, so the nail force is more mobilized at the lower part. When there
is local loading applied on the slope, the nail force of the top row at the centre of the
model is much more mobilized (loading is applied at the range of the model center). For
slope with local loading at the model center (Figures 6.3 to 6.5), the nail force
distribution at the model edge has no great difference from the slope with no loading
(the bottom row is more mobilized than the top row), since at the edge of the model, it
is far away from the loading which is applied at the center of the model and the
influence of the loading is small. Around the center of the model, when the loading
length is small (L/B=1, Figure 6.3), the nail forces of the top and middle rows are more
mobilized, but the nail force of the bottom row is nearly not affected. When the loading
length is large (L/B=2 and 4, Figures 6.4 and 6.5), the nail forces of the top and middle
rows are much more mobilized and the nail force of the bottom row is also slightly
more mobilized.

Chapter 6: Effects of Curvature and Local Loading on 3D Soil Nailed Slope

- 144 -
6.3 Numerical simulation for idealized convex and concave slopes
In this section, several idealized concave and convex slope models are developed to
investigate the effect of curvature on soil nailed slope. The slope height is 10m and
slope angle is 45 with properties of density=19.5kN/m3, c=10kPa and =36 (same as
the model in chapter 4, Figure 4.19). In this analysis, two different groups of slope
models are developed. In the first group, 10m length soil nails are installed at 1.5m
centers horizontally and 2m vertically and no external loading is applied on the slope. In
the second group, 10m length soil nails are installed at 1.5m centers horizontally and
2m vertically together with a 4m width 100kPa external loading applied on the crest of
the slope. In each group, five models are developed (two concave models with 17m and
35m radius at the slope middle height, two convex models with 17m and 35m radius at
the slope middle height and one plane model with no curvature). The typical slope
section is shown in Figure 6.7. Only a slice mode with a 1.5m thickness at the middle of
the slope is adopted in the 3D analysis, and the typical models are shown in Figure 6.8.

The variation of the safety factors with different curvatures are shown in Tables 6.1 to
6.2. For concave slope, the effect of the curvature is obvious and the FOS increases with
the decrease of radius. For convex slope, it seems that the curvature does not have
distinct effect on the factor of safety which is similar to the un-reinforced slopes
investigated in chapter 4.

The slip surfaces for different curvature with no loading is applied are shown in Figure
6.9. As discussed in chapter 4, when the slope is not reinforced, the curvature has nearly
no influence on the location of slip surface. When the soil nail is included in the slope
(Figure 6.9), it can be seen that the slip surface for reinforced slope is influenced by the
curvature, particularly for concave slope. When the radius is smaller for concave slope,
Chapter 6: Effects of Curvature and Local Loading on 3D Soil Nailed Slope

- 145 -
the slip surface will be shallower. For convex slope with soil nails, the curvature has no
major effect on the slip surface under normal curvature.

6.4 Numerical simulation for intersected slopes
In this section, two different intersected slope models are developed. For the concave
slope model, the intersection angle is 135 and for the convex slope the intersection
angle is 145. The slope height is 6 m and the slope angle is 45 (same as the model in
chapter 4, Figure 4.23).

Firstly, no loading is applied on the slope and the results are shown in Figure 6.10. For
soil nailed slope, the effect of curvature for convex slope is still not obvious, and the
FOS is nearly the same as that for a plane slope, and the effect of curvature is more
obvious for concave slope with a larger FOS. Secondly, rectangular shaped vertical
loading is applied on the soil nailed slope top. The loading length is 4m and the loading
width B is 2m, while the edge of the loading is 1m away from the crest of the slope. For
the locally loaded soil nailed slope, two different models are developed. In the first
model, a very good nail head is simulated by a thin elastic plate and the nail is fixed on
the nail head. There is no nail head failure in this model failure and the FOS and slip
surface for different curvature are shown in Figure 6.11. In the second model, no nail
head is simulated, so only the bond stress within the soil mass is mobilized and the FOS
and slip surface for different curvature are shown in Figure 6.12. For the first model, the
FOS has an obvious increase (Figure 6.11) compared with that for an unreinforced slope
(Figure 4.24). For the second model which is a typical face failure, the increasing of
FOS is very small (Figure 6.12) compared with an unreinforced slope (Figure 4.24).
This example shows that a good nail head design is very important in practice. In both
Chapter 6: Effects of Curvature and Local Loading on 3D Soil Nailed Slope

- 146 -
the models, the effect of the curvature is not obvious for convex slope but is more
obvious for concave slope.

The nail maximum tension force distribution is shown in Figures 6.13 to 6.19. For slope
with no loading, the nail force of the bottom row is more mobilized (Figures 6.13 and
6.14). For convex slope, the nail force in a certain row is smaller at the slope
intersection area, since in this area the nail density is a slightly larger because of the nail
overlapping from both sides of the slope. For slope with loading (Figures 6.15 to 6.19),
the nail force in the top and middle rows is much more mobilized, especially for the
face failure situation in which the nail head is not simulated.

For un-reinforced slopes, it has been shown in chapter 4 that the curvature has little
influence on the stability of normal convex slopes, but has distinct effect on the steep
convex slope with local loading. In this chapter, it is shown in above that the effect of
curvature can also be ignored for normal convex slope reinforced with nail. In order to
examine the influence of curvature on steep convex slope with soil nails, the vertical cut
model analyzed in chapter 4 (Figure 4.26) is investigated again, and the soil nails are
installed in the model. The slope geometry and applied loading is the same as Figure
4.26. The nails are installed at 1.5m centers horizontally and vertically. Two different
situations are considered. In the first situation, the nail head is simulated by thin elastic
plate, and the failure is pullout failure (Figure 6.20). In the second situation, no nail
head is simulated, and the failure is face failure (Figure 6.21). Compared with the un-
reinforced slope (Figure 4.26), the factor of safety has an obvious increase in the first
situation and has only a slight increase in the second situation. Both situations show that
the curvature has distinct effect on the stability of this locally loaded vertical cut slope
reinforced with nails, and the factor of safety is smaller for the convex slope model.
Chapter 6: Effects of Curvature and Local Loading on 3D Soil Nailed Slope

- 147 -
Based on the above analysis, the intersected slope reinforcement design should be as
follows. For the concave intersection part, less reinforcement can be used as the effect
of curvature can improve the slope stability. For the convex intersection part, if the
slope is not steep, the reinforcement design can be the same as a plane slope, but if the
slope is steep, more reinforcement should be used. When soil nail is used to reinforce
intersected slope, even though the same soil nail design is used at the convex part, since
there is some overlapping of the nail from the two intersected faces, the actual nail
density at this part is slightly larger which will enhance the slope stability, but the
improvement should be determined by a detailed analysis according to the situation in
practice.

6.5 Some small scale model testing results
In recent years, some small scale tests are conducted in the laboratory of the Hong Kong
Polytechnic University to investigate the effect of curvature and the soil nail
performance. Three model tests have been conducted by Tsui (2007), and the results of
these tests can be used to illustrate some of the numerical results in this chapter. These
model slopes are relatively steep with a 65 slope angle and slope height is 0.8m. The
first test is about a soil nailed intersected convex slope. The second test is a soil nailed
slope with no curvature. The third test is a plane slope with no soil nail. In all these
tests, the load is applied with a 130 mm offset from crest of the slope, and the load is
increased gradually until the slope fails.

For the two soil nailed slopes, a good nail head is not constructed, so the soil nail is not
pulled out and a face failure is mobilized where the bond stress within the soil mass is
effective in the stabilization. The slope failure surfaces for different models are shown
in Figure 6.22. For the weight of the failure mass, it is about 125kg in the Test 1, 194kg
Chapter 6: Effects of Curvature and Local Loading on 3D Soil Nailed Slope

- 148 -
in the Test 2 while it is about 382kg in the Test 3. For the maximum load, it is about
26kN in Test 1, 44.2kN in Test 2 while it is 35kN in Test 3.

For soil nailed slope, the maximum load for the convex slope is smaller than the one for
the plane slope, and the sliding mass is also smaller. This shows that for steep slope, a
convex slope is less safe than a plane slope under local loading which verifies the
numerical simulation results of the vertical cut slope presented in the previous section.
For the two slopes with no curvature, the maximum load for soil nailed slope is larger
than the one for unreinfirced slope, but the sliding mass is smaller. It means that the soil
nail has improved the safety of the slope.

6.6 Discussion and Conclusion
For a soil nailed slope without external load, the nail force of the bottom row is usually
more mobilized at the limit state. For locally loaded soil nailed slopes, the nail force of
the top row gets more mobilized as the movement of the top part of the slope is larger.

For the idealized curvature slope, it has been shown in chapter 4 that the curvature has
no major effect on the location of the slip surface of unreinforced slope. When soil nail
is included in the slope, the slip surface is influenced by the curvature, especially for
concave slope. When the radius is smaller for concave slope, the slip surface is
shallower.

The curvature does not have any noticeable effect on the factor of safety for normal
convex slopes. When local loading is applied on the top of convex slope and if the slope
is steep, the effect of curvature will be obvious and the FOS of convex slope gets
smaller, and this phenomenon is similar with the result for un-reinforced slopes in
Chapter 6: Effects of Curvature and Local Loading on 3D Soil Nailed Slope

- 149 -
chapter 4. The present analysis also shows that a good nail head design is very
important in practice.

Based on the above analysis, the intersected slope reinforcement design should be as
follows. For the concave intersection part, less reinforcement can be used as the effect
of curvature can improve the slope stability. For the convex intersection part, if the
slope is not steep, the reinforcement design can be the same as that for an unreinforced
slope. If the slope is steep, more reinforcement should be used. When soil nail is used to
reinforce intersected slope, even though the same soil nail design is used at the convex
part, since there is some overlapping of the nail from the two intersected faces, the
actual nail density at this part is slightly larger which will further enhance the slope
stability, but the actual contribution to the slope stability should be determined by a
detailed analysis according to the situation in practice.

The physical model tests also show that for a steep slope, the convex slope is less safe
than a plane slope under local loading.










Chapter 6: Effects of Curvature and Local Loading on 3D Soil Nailed Slope

- 150 -

Table 6.1 Variation of FOS respect to curvature (nail 10m, no load)
Radius of curvature (m) 17 35 (plane)
Concave(SRM) 2.48 2.14 1.92
Convex(SRM) 1.97 1.92 1.92


Table 6.2 Variation of FOS respect to curvature (nail 10m, with 4m width load 100kPa)
Radius of curvature (m) 17 35 (plane)
Concave(SRM) 1.57 1.46 1.39
Convex(SRM) 1.47 1.40 1.39


















Chapter 6: Effects of Curvature and Local Loading on 3D Soil Nailed Slope

- 151 -

45
6m
1m 2m
Cohesion =20kN/m
2

Friction angle =20
Density =20kN/m
3

q=200kPa
1.5m
Soil nail, length=8m

Figure 6.1 The geometry of the slope under local loading


(a) L/B=1, FOS=1.61 (b) L/B=2, FOS=1.36 (c) L/B=4, FOS=1.17
Figure 6.2 The slip surface for different loading length when B=2m (nail 8m,
load=200kPa)

0
5
10
15
20
25
30
35
40
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
L (m)
m
a
x
i
m
u
m

n
a
i
l

f
o
r
c
e

(
k
N
)
top
middle
bottom

Figure 6.3 Nail maximum tension force distribution for L/B=1

Chapter 6: Effects of Curvature and Local Loading on 3D Soil Nailed Slope

- 152 -
0
5
10
15
20
25
30
35
40
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
L (m)
m
a
x
i
m
u
m

n
a
i
l

f
o
r
c
e

(
k
N
)
top
middle
bottom

Figure 6.4 Nail maximum tension force distribution for L/B=2

0
5
10
15
20
25
30
35
40
45
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
L (m)
m
a
x
i
m
u
m

n
a
i
l

f
o
r
c
e

(
k
N
)
top
middle
bottom

Figure 6.5 Nail maximum tension force distribution for L/B=4

0
5
10
15
20
25
30
35
0 2 4 6 8 10 12 14 16 18 20
L (m)
m
a
x
i
m
u
m

n
a
i
l

f
o
r
c
e

(
k
N
)
top
middle
bottom

Figure 6.6 Nail maximum tension force distribution with no loading (FOS=1.94)
Chapter 6: Effects of Curvature and Local Loading on 3D Soil Nailed Slope

- 153 -


10m
4m
q=100kPa
nail, length=10m
45
2.0m

Figure 6.7 The geometry of the slope section



(a) convex (b) concave (c) plane
Figure 6.8 Mesh plot of concave and convex models reinforced with nails



(a) concave, r=17m (b) concave, r=35m (c) plane, r=

(d) convex, r=35m (e) convex, r=17m
Figure 6.9 Slip surfaces at different curvatures (with nail, no load)


Chapter 6: Effects of Curvature and Local Loading on 3D Soil Nailed Slope

- 154 -


(a) 135 concave, FOS=2.08 (b) no curvature, FOS=1.94 (c) 145 convex, FOS=1.93
Figure 6.10 Slip surfaces for different curvatures with 200kPa loading (no load, nail
8m)



(a) 135 concave, FOS=1.59 (b) no curvature, FOS=1.36 (c) 145 convex, FOS=1.40
Figure 6.11 Slip surfaces for different curvatures with 200kPa loading (nail 8m, pullout
failure)



(a) 135 concave, FOS=1.39 (b) no curvature, FOS=1.23 (c) 145 convex, FOS=1.27
Figure 6.12 Slip surfaces for different curvatures with 200kPa loading (nail 8m, face
failure)


Chapter 6: Effects of Curvature and Local Loading on 3D Soil Nailed Slope

- 155 -
0
5
10
15
20
25
30
35
40
45
50
-8 -6 -4 -2 0 2 4 6 8
L (m)
m
a
x
i
m
u
m

n
a
i
l

f
o
r
c
e

(
k
N
)
top
middle
bottom

Figure 6.13 Nail maximum tension force distribution for a 135 concave slope with no
loading

0
5
10
15
20
25
30
35
40
45
-8 -6 -4 -2 0 2 4 6 8
L (m)
m
a
x
i
m
u
m

n
a
i
l

f
o
r
c
e

(
k
N
)
top
middle
bottom

Figure 6.14 Nail maximum tension force distribution for a 145 convex slope with no
loading

0
10
20
30
40
50
60
-8 -6 -4 -2 0 2 4 6 8
L (m)
m
a
x
i
m
u
m

n
a
i
l

f
o
r
c
e

(
k
N
)
top
middle
bottom

Figure 6.15 Nail maximum tension force distribution for a 135 concave slope with
200kPa loading
Chapter 6: Effects of Curvature and Local Loading on 3D Soil Nailed Slope

- 156 -
0
5
10
15
20
25
30
35
40
45
-8 -6 -4 -2 0 2 4 6 8
L (m)
m
a
x
i
m
u
m

n
a
i
l

f
o
r
c
e

(
k
N
)
top
middle
bottom

Figure 6.16 Nail maximum tension force distribution for a 145 convex slope with
200kPa loading

0
5
10
15
20
25
-8 -6 -4 -2 0 2 4 6 8
L (m)
m
a
x
i
m
u
m

n
a
i
l

f
o
r
c
e

(
k
N
)
top
middle
bottom

Figure 6.17 Nail maximum tension force distribution for a 135 concave slope with
200kPa loading (face failure)

0
5
10
15
20
25
-8 -6 -4 -2 0 2 4 6 8
L (m)
m
a
x
i
m
u
m

n
a
i
l

f
o
r
c
e

(
k
N
)
top
middle
bottom

Figure 6.18 Nail maximum tension force distribution for a 145 convex slope with
200kpa loading (face failure)
Chapter 6: Effects of Curvature and Local Loading on 3D Soil Nailed Slope

- 157 -


0
5
10
15
20
25
0 2 4 6 8 10 12 14 16 18 20
L (m)
m
a
x
i
m
u
m

n
a
i
l

f
o
r
c
e

(
k
N
)
top
middle
bottom

Figure 6.19 Nail maximum tension force distribution for a no curvature slope with
200kpa loading (face failure)




(a) convex, FOS=1.61 (b) plane, FOS=1.79
Figure 6.20 Slip surface of vertical cut slope with 3m long and 3m wide 50kPa loading
(pullout failure)

Chapter 6: Effects of Curvature and Local Loading on 3D Soil Nailed Slope

- 158 -

(a) convex, FOS=1.01 (b) plane, FOS=1.33
Figure 6.21 Slip surface of vertical cut slope with 3m long and 3m wide 50kPa loading
(face failure)



(a) test 1, soil nailed convex slope

(b) test 2, soil nailed no curvature slope (c) test 3, non soil nailed no curvature slope
Figure 6.22 Different failure shapes after failure mass was removed (after Tsui, 2007)
Chapter 7: Strength Reduction Analysis for Slope Reinforced with One Row of Piles

- 159 -

CHAPTER 7: STRENGTH REDUCTION ANALYSIS FOR
SLOPE REINFORCED WITH ONE ROW OF PILES

7.1 Introduction
It is an effective solution to increase slope stability by installing piles. The piles used in
slope stabilization are considered as passive piles because they are usually subjected to
lateral force arising from the horizontal movements of the surrounding soil. The
interaction between the passive piles and the soil is complicated, as the lateral forces
exerting on the piles are related to the soil movements which are in turn influenced by
the presence of the piles. A number of methods have been proposed to evaluate the
performance and design of the piled slopes. These methods include limit equilibrium
method, limit analysis method, strength reduction method, and others. Due to the
complicated interaction between soil and pile, even though some limit equilibrium
methods are employed for piled slope analysis, they are still not widely accepted by
engineers. In this study, the strength reduction method will hence be used to conduct the
piled slope analysis by software FLAC3D.

The shear strength reduction method has been used by Cai and Ugai (2000) for the
analysis of slopes reinforced by one row of piles. Won et al. (2005) have also analyzed
the same slope by Cai and Ugai (2000) using the three-dimensional finite difference
code FLAC3D by the SRM. In the piled slope analysis by Won et al. (2005) and Cai
and Ugai (2000), the location of the slip surface was determine by the maximum shear
force in the pile so that a very deep slip surface was determined while the maximum
shear strain in soil was not considered. In the present study, slip surface will be
Chapter 7: Strength Reduction Analysis for Slope Reinforced with One Row of Piles

- 160 -
determined by the shear strain rate and the failure mode is determined with respect to
different pile spacing. It is found that the pile spacing has a major influence on the
failure mode, and a more realistic slip surface is found from the present study. The
upper and lower bounds of the factor of safety and the optimum pile location will also
be discussed in the present study.

7.2 Failure mode of slopes with different pile spacing
For the slope considered by Won et al. (2005) and Cai and Ugai (2000), it is 10m in
height with a gradient of 1V:1.5H (Figure 7.1). Two symmetric extreme boundaries are
used so that the problem consists of a row of piles with a plane of symmetry. Steel tube
pile with an outer diameter (D) of 0.8m is used in this study. The piles are treated as a
linear elastic solid material and are installed in the middle of the slope with L
x
=7.5m
and center-to-center spacing (s) varies from 2D to 8D. The piles are embedded and
fixed into the bedrock or a stable layer (infinite pile length assumption). In this model,
the pile head is free. The cohesive strength, friction angle, elastic modulus, Poisson ratio
and density of the soil are 10kPa, 20, 200MPa, 0.25, and 20kN/m
3
respectively. The
elastic modulus and Poisson ratio of the piles are 60000MPa and 0.2 respectively.

When the slope is not reinforced with piles, the factor of safety by the SRM is 1.20 and
the slip surface is shown in Figure 7.2. The slip surfaces and the factors of safety for
different pile spacing are shown in Figures 7.3 to 7.8. It is reasonable that the factor of
safety decreases with the increase of pile spacing. In Figures 7.3 to 7.8, the slip surfaces
through different sections are shown (section y=0 is the plane through pile centerline;
section y=0.4 is the plane tangent to the pile outer surface; and for the section through
soil midway between piles, y varies from 0.8 to 3.2 with respect to pile spacing varies
from 2D to 8D). When s=2D (Figure 7.3), the slip surface is clearly divided into two
Chapter 7: Strength Reduction Analysis for Slope Reinforced with One Row of Piles

- 161 -
parts even at the section through the soil midway between the piles. When the section
varies from the plane through the pile centerline to the plane through the soil midway
between the piles, the shear strain in the soil gets more mobilized. When s=3D (Figure
7.4), the slip surface near the pile is still clearly divided into two parts. The slip surfaces
at the section through the soil midway between the piles are slightly connected, but the
overall slip surface is still nearly divided into two parts. When the pile spacing increases
from 4D to 6D (Figures 7.5 to 7.7), the slip surface at the sections near the pile is also
divided into two parts due to the presence of the piles, but the two parts of slip surfaces
gets deeper near the pile location, so there is a clear shear strain mobilization at the
interface of the pile-soil in the vertical direction. The slip surface at the section through
the soil midway between the piles gets more connected and deeper, and at the
connection of the two parts of the slip surfaces, there is clear shear strain mobilization in
the vertical direction. When the pile spacing increases to 8D (Figure 7.8), the slip
surface at the section through the soil midway between the piles is clearly one single
slip surface, though the slip surface at the sections near the pile is still divided into two
parts due to the presence of the piles. For a section which is not far from the pile, for
example y=1.0 (Figure 7.8d), a clear shear strain mobilization in the vertical direction is
also found at the pile location. When the section is far enough from the pile, for
example y=1.8 (Figure 7.8e), the vertical shear strain mobilization at the slip surface
center disappears.

The slip surfaces at the section through the soil midway between the piles are compared
for different pile spacing in Figure 7.9 (slip surface can be seen easily from the shear
strain mobilization and are also shown in dashed lines). The slip surface for the slope
with no pile is shown by the solid line. When the pile spacing is small, the slip surface
is shallow and is nearly divided into two parts. With the increase in pile spacing, the slip
Chapter 7: Strength Reduction Analysis for Slope Reinforced with One Row of Piles

- 162 -
surface gets deeper and the two parts of the slip surface gets more connected. When the
pile spacing is large enough, the two parts of the slip surface gradually merge to a clear
single slip surface which is nearly the same as the slip surface for unreinforced slope.
This means that the slip surface of a piled slope is usually shallower than that with no
pile, and this result is totally different from that by Won et al. (2005) and Cai and Ugai
(2000) who obtained a very deep slip surface for piled slopes based on the maximum
shear force in the pile (Figure 7.10). The shear force distribution for pile spacing equal
to 3D is shown in Figure 7.11. It is clearly seen that the location of the maximum shear
force is very deep and is far away from the real slip surface, therefore, the maximum
shear force location is not necessarily the location of the critical slip surface for a piled
slope.

7.3 Upper and lower bound of the factor of safety
Identifying the upper and lower bounds of the factor of safety is very useful and
important before the detailed design of a pile supported slope. The factor of safety for
slopes with no pile can be taken as the lower bound, while the FOS for a slope with a
pile wall can be taken as the upper bound. For the slope model analyzed in the previous
section, the lower and upper bounds of the FOS are 1.20 (Figure 7.2) and 1.89 (Figure
7.12) respectively. The FOS for different pile spacing is shown in Table 7.1.

Based on the above study, a very simple approach for the piled slope SRM analysis is
proposed before conducting a detailed design. Firstly, a slope with no pile and with a
pile wall is analyzed to determine the lower and upper bounds of the factor of safety.
For these two cases, only a 2D analysis will be required and this will save significant
time and effort. Secondly, the upper bound solution is actually equivalent to the case of
s=D, while the lower bound is equivalent to s=. The general curve of the safety factor
Chapter 7: Strength Reduction Analysis for Slope Reinforced with One Row of Piles

- 163 -
variation with respect to spacing is shown in Figure 7.13a. When s/D increases from 1
to infinity, the factor of safety will decrease from the upper bound to the lower bound.
The FOS usually decreases quicker when the spacing is relatively small and becomes
gradual when the spacing is large. In practice, the pile spacing will not be very large,
therefore, we only need to consider the range for relatively small spacing. Although the
decrease of the FOS from the upper bound to the lower bound is not easy to predict for
a general problem, at the preliminary design, we can assume the FOS decreases linearly
from the upper bound to the lower bound when the spacing increases from 1D to X
times D (shown in Figure 7.13a). For the example analyzed above, when X equals to 10
or 12, a good approximate prediction can be obtained by a simple interpolation (the
result is shown in Figure 7.13b for X=12). The X value in the range of 10 to 14 is
recommended after a serious of different case studies. According to the proposed
approach, an approximate pile spacing can be determined and fine tune of the spacing
by a detailed 3D analysis can be conducted afterwards.

7.4 Optimization of the pile position
In piled slopes, the position of the pile is very important and its effect has been
discussed by Won et al. (2005) and Cai and Ugai (2000), who came to the same
conclusion that the improvement of the slope stability will be largest when the piles are
located in the middle of the slopes. In general, the optimal position should also be
dependent on the s/D ratio, but it is found that the optimum pile location is not sensitive
to this ratio based on a series of trials. For simplicity, the situation of a pile wall (s=D)
will be used to conduct the pile position optimization since this 2D analysis is a time
saving analysis (3D SRM is extremely time consuming in the analysis and some
problems takes 2-3 weeks by a Pentium 4 for the analysis). As shown in Figure 7.12, if
a pile wall is installed in the middle of the slope, the slip surface is nearly divided into
Chapter 7: Strength Reduction Analysis for Slope Reinforced with One Row of Piles

- 164 -
two equal parts (actually when the pile spacing is very small, for example s=2D, the slip
surface is also divided into two parts). If the two equal half do not interfere with each
other, the two local factors of safety will be identical and the overall factor of safety will
be the optimum as proposed by Won et al. (2005) and Cai and Ugai (2000). It is
however found that the two half of soil mass interfere with each other so that there is
some small difference between the lower and upper parts. As shown in Figure 7.12, the
slip surface is divided into two parts, but the shear strain in the upper part is slightly
more mobilized at the limit state. This means that the factor of safety for the upper part
is lower than that for the lower part, and the overall FOS of the slope will be controlled
by the upper part of the slip surface. In order to compare the safety of the lower and
upper parts, two additional models are developed. In the first model, the shear strength
of the soil at the upper part of the pile is increased to a very large value, so slope failure
will occurs only at the lower part and the FOS is 1.94 (Figure 7.14). In the second
model, the shear strength of the soil at the lower part of the pile is increased to a very
large value, so slope failure will occur only at the upper part and the FOS is 1.87
(Figure 7.15). It can be seen that the FOS of the lower part is a little larger than the
upper part, but the difference is not large. If the pile position is 0.2m above the middle
of the slope, the factors of safety of the lower part and the upper part are both 1.91
(Figure 7.16) which will be the optimal solution. This means that the optimal pile
position for this slope model is 0.2m above the middle of the slope.

For the optimal location of the piles, Poulos (1995) has proposed another guideline
where the piles are installed around the center of the critical slip surface to avoid
shifting the slip surface either in front of or behind the piles. For the slope model
analyzed above, the slip surface with no pile is shown in Figure 7.17 and it is compared
with the optimized pile location. It can be seen that the exact optimized pile location lies
Chapter 7: Strength Reduction Analysis for Slope Reinforced with One Row of Piles

- 165 -
between the middle of the slope and the middle of the slip surface of the slope with no
pile, but it is much closer to the middle of the slope.

In order to get more insight into this issue, another slope model with different soil
property is analyzed. The soil cohesion and friction angle are 20kPa and 10 degree
respectively for this model. The FOS is 1.14 for this slope with no pile and the slip
surface (Figure 7.18) is much deeper than the previous slope model. For this model, if
the pile is installed at the middle of the slope, the FOS is 2.22 and the slip surface is
shown in Figure 7.19a. It can be seen that obvious shear strain is mobilized only at the
upper part. This means the lower part is much safer than the upper part which is also
verified by Figure 7.19b (FOS is 2.48 for the lower part) and Figure 7.19c (FOS is 2.16
for the upper part). If the pile is installed 0.55m towards the crest of the slope (Figure
7.20), the FOS of the lower part (2.32) is nearly the same as the upper part (2.33). The
optimized pile position is shown in Figure 7.21, which also lies between the middle of
the slope and the middle of the slip surface of the slope with no pile, but it is still much
closer to the middle of the slope. The optimal pile position for the above two models are
both at some distance towards the crest of the slope as measured from the middle of the
slope, and it is interesting that this distance for both models are nearly one sixth of the
interval between the center of the slope and the center of the slip surface.

Based on the above analysis, the main conclusions are as follows. For slopes reinforced
with one row of piles, the optimal pile position lies between the middle of slope and the
middle of the slip surface of the slope with no pile, and is very close to the middle of the
slope. For slopes with sandy soil (soil cohesion is small but friction angle is large), the
slip surface is shallow and the upper end of the slip surface is very close to the slope
crest. The middle of the slip surface will also be close to the middle of the slope, and the
Chapter 7: Strength Reduction Analysis for Slope Reinforced with One Row of Piles

- 166 -
optimal pile position is nearly located at the middle of the sandy soil slope. For slopes
with clayey soil (soil cohesion is large but friction angle is small), the slip surface is
deep and the upper end of the slip surface is far away from the slope crest, so the middle
of the slip surface is also far away from the middle of the slope while the optimal pile
position will be slightly more away from the middle of the slope.

7.5 Conclusion
For slopes reinforced with one row of piles, when the pile spacing is small, the slip
surface is shallow and is nearly divided into two parts. With the increase in the pile
spacing, the slip surface becomes deeper and the two parts of slip surface become more
connected. When the pile spacing is large enough, the two parts of the slip surface
gradually turn into a clear single slip surface which is nearly the same as the slip surface
with no pile. This means that the slip surface of the piled slope is usually shallower than
that of a slope with no pile. This result is totally different from the results based on the
maximum point of shear force by which a very deep slip surface for the piled slope is
usually obtained. The present study has revealed that the maximum point of shear force
is very deep and is far away from the real slip surface, therefore, the pile maximum
shear force location is not necessarily the location of the critical slip surface of a piled
slope.

It is extremely time-consuming to carry out a detailed 3D piled slope analysis, and lots
of time has been spent in carrying out the analysis for the present study. For practical
analysis and design, a useful guideline to the engineers has been proposed in this study.
The upper and lower bounds of the factor of safety can be determined easily by a 2D
analysis and they are useful for the preliminary design before conducting a detailed 3D
Chapter 7: Strength Reduction Analysis for Slope Reinforced with One Row of Piles

- 167 -
analysis and design. Based on the guideline as shown in Figure 7.13, an economical
design can be obtained without tedious efforts in finding an optimal pile spacing.

The optimal pile position for a slope reinforced with one row of piles is found to lie
between the middle of slope and the middle of slip surface of the slope with no pile. The
precise location is usually very closer to the middle of the slope. For slopes with sandy
soil, the optimal pile position will be nearly located at the middle of the slope so that the
middle of the slope can be chosen for the optimum design. For slope with clayey soil,
the exact optimal pile position is usually located slightly towards the crest of the slope
instead of being located at the middle of the slope.
















Chapter 7: Strength Reduction Analysis for Slope Reinforced with One Row of Piles

- 168 -

Table 7.1 Factor of safety with different pile spacing
pile
spacing
pile wall
(s=D)
s=2D s=3D s=4D s=5D s=6D s=8D
No pile
(s=)
FOS
1.89
(upper
bound)
1.78 1.72 1.61 1.55 1.52 1.42
1.20
(lower
bound)





















Chapter 7: Strength Reduction Analysis for Slope Reinforced with One Row of Piles

- 169 -


10.0m
15.0m
10.0m
20.0m
10.0m
X
Y
Z
Lx
D
s
s

Figure 7.1 Slope model and finite difference mesh


Figure 7.2 Slip surface of the slope with no pile (FOS=1.20)


(a) y=0, through pile centerline (b) y=0.2 (c) y=0.4

(d) y=0.6 (e) y=0.8, through soil midway between piles
Figure 7.3 Slip surface at different sections for s=2D (FOS=1.78)

Chapter 7: Strength Reduction Analysis for Slope Reinforced with One Row of Piles

- 170 -



(a) y=0, through pile centerline (b) y=0.2 (c) y=0.4

(d) y=0.6 (e) y=0.8 (f) y=1.2, midway through soil
Figure 7.4 Slip surface at different sections for s=3D (FOS=1.72)



(a) y=0, through pile centerline (b) y=0.2 (c) y=0.4

(d) y=0.8 (e) y=1.2 (f) y=1.6, midway through soil
Figure 7.5 Slip surface at different sections for s=4D (FOS=1.61)



Chapter 7: Strength Reduction Analysis for Slope Reinforced with One Row of Piles

- 171 -



(a) y=0, through pile centerline (b) y=0.2 (c) y=0.4

(d) y=0.8 (e) y=1.4 (f) y=2.0, midway through soil
Figure 7.6 Slip surface at different sections for s=5D (FOS=1.55)



(a) y=0, through pile centerline (b) y=0.2 (c) y=0.4

(d) y=1.0 (e) y=1.6 (f) y=2.4, midway through soil
Figure 7.7 Slip surface at different sections for s=6D (FOS=1.52)



Chapter 7: Strength Reduction Analysis for Slope Reinforced with One Row of Piles

- 172 -



(a) y=0, through pile centerline (b) y=0.2 (c) y=0.4

(d) y=1.0 (e) y=1.8 (f) y=3.2, midway through soil
Figure 7.8 Slip surface at different sections for s=8D (FOS=1.42)



(a) s=2D (b) s=3D

(c) s=4D (d) s=5D

(e) s=6D (f) s=8D
Figure 7.9 Slip surface at the section of soil midway between piles

Chapter 7: Strength Reduction Analysis for Slope Reinforced with One Row of Piles

- 173 -


Flac 3D(6.33m)
extremepoint
of shear force
Bishop(slip depth: 3.6m)
Pileelement detailed


Critical
slip circle
Extreme point
of shear force
D1/D=2.0

(a) Won et al. (2005) (b) Cai and Ugai (2000)
Figure 7.10 Slip surface obtained by SRM based on extreme point of shear force and
comparison with critical slip circle obtained by Bishops simplified method


Extreme point
of shear force

Figure 7.11 Shear force distribution of the piled slope for s=3D



(a) mesh plot (b) slip surface
Figure 7.12 Slip surface and mesh for the slope with pile wall (FOS=1.89)

Chapter 7: Strength Reduction Analysis for Slope Reinforced with One Row of Piles

- 174 -

s/D
F
a
c
t
o
r

o
f

s
a
f
e
t
y

1
Upper bound of FOS
Lower bound of FOS
X

1
1.1
1.2
1.3
1.4
1.5
1.6
1.7
1.8
1.9
2
0 1 2 3 4 5 6 7 8 9 10 11 12 13
s/D
f
a
c
t
o
r

o
f

s
a
f
e
t
y
Upper bound
Lower bound

(a) general curve (b) X=12
Figure 7.13 Factor of safety with respect to different pile spacing



(a) mesh plot (b) slip surface
Figure 7.14 Slip surface and mesh for the slope with strong soil at upper part
(FOS=1.94)



(a) mesh plot (b) slip surface
Figure 7.15 Slip surface and mesh for the slope with strong soil at lower part
(FOS=1.87)

Chapter 7: Strength Reduction Analysis for Slope Reinforced with One Row of Piles

- 175 -


(a) pile wall, FOS=1.88 (b) strong soil up part, FOS=1.91 (c) strong soil low part, FOS=1.91
Figure 7.16 Slip surface for the slope with pile wall installed at 0.2m towards the slope
crest as measured from the middle of slope

X
Z
Middle of the slope
(X=17.5)
Middle of the slope slip surface
with no pile (X=18.7)
Optimized pile position
(X=17.7)
Slip surface
with no pile
0, 0
0, 10
35, 0
35, 20
10, 10
25, 20

Figure 7.17 Pile position for the slope with soil cohesion 10kpa and friction angle 20
degree


Figure 7.18 Slip surface of the slope with no pile for soil cohesion=20kpa and friction
angle 10 degree (FOS=1.14)


Chapter 7: Strength Reduction Analysis for Slope Reinforced with One Row of Piles

- 176 -


(a) pile wall, FOS=2.22 (b) strong soil up part, FOS=2.48 (c) strong soil low part, FOS=2.16
Figure 7.19 Slip surface for the slope with pile installed in the middle of slope (soil
cohesion=20kpa, friction angle=10 degree)


(a) pile wall, FOS=2.33 (b) strong soil up part, FOS=2.32 (c) strong soil low part, FOS=2.33
Figure 7.20 Slip surface for the slope with pile installed 0.55m upper of the middle of
slope (soil cohesion=20kpa, friction angle=10 degree)

X
Z
Middle of the slope
(X=17.5)
Middle of the slope slip surface
with no pile (X=20.625)
Optimized pile position
(X=18.05)
Slip surface
with no pile
0, 0
0, 10
35, 0
35, 20
10, 10
25, 20

Figure 7.21 Pile position for the slope with soil cohesion 20kpa and friction angle 10
degree
Chapter 8: Stability Analysis of Slope with Water Flow by Strength Reduction Method

- 177 -

CHAPTER 8: STABILITY ANALYSIS OF SLOPE WITH
WATER FLOW BY STRENGTH REDUCTION METHOD

8.1 Introduction
Pore water pressure is one of the major reasons for slope instability. In limit equilibrium
method which is widely used in engineering practice, the most common way in defining
the pore-water pressure is the piezometric line, which is actually a hydrostatic condition.
Since the assumption of a hydrostatic pore-water pressure is strictly not correct for most
seepage situations, a correction factor is used in some software. Furthermore, if the
pore-water pressure distribution is known by finite element or finite difference analysis,
the LEM can also be integrated with the pore water pressure from the seepage analysis
in the stability calculation. The SRM can also adopt the results from the seepage
analysis easily, and the mesh for the SRM and seepage analysis can actually be the
same.

As mentioned in chapter 2, for slope stability analysis with water flow, the way to
include seepage forces in the LEM has caused various confusions, and some researchers
(Greenwood, 1983, 1985; King, 1989) have introduced some effective-stress methods of
slices which include the interslice water forces. These approaches are however more
complicated in the analysis and are not adopted in commercial programs. Greenwood
(1983) and Duncan and Wright (2005) have shown that there are cases where the
refined approaches may have noticeable differences with the classical methods. If the
SRM is used, the pore-pressure will affect the effective stress for which the stability
Chapter 8: Stability Analysis of Slope with Water Flow by Strength Reduction Method

- 178 -
analysis is based on. The confusions about the effect of water in the LEM will not
appear in the SRM.

In this chapter, two and three-dimensional strength reduction analyses will be carried
out by software FLAC3D to study the effects of water flow on slope under several
cases. In the following examples, the main focus is the steady state flow situation while
the transient flow effect is not considered, so the value of the soil permeability has no
influence on the SRM results. In order to reduce the running time, a relatively large
value is assigned to the soil permeability (0.01cm/s). This study has demonstrated that
besides the lowering of the factor of safety, the failure mechanism may also be affected
by the seepage flow. Many engineers view that densely populated soil nails may affect
seepage flow, but it is demonstrated in this study that for practical purposes, the effect
of soil nail on the seepage flow can be neglected.

8.2 Stability analysis for a simple slope with seepage flow
In this section, a two-dimensional 6m height slope with 45 slope angle is analyzed. A
10m height model is developed in which the water is 4m height on the left and 10m
height on the right side. The pore-pressure and the flow vector distribution for the free-
surface seepage flow analysis are shown in Figure 8.1. For different soil properties, the
FOS and slip surfaces are shown in Figures 8.2 to 8.5, and the results are compared with
the case with no water. Obviously, with the seepage flow, the FOS is much smaller than
the corresponding case without water. For sandy soil, the decrease of the FOS with
seepage flow is larger than that for the clayey soil. This means that a sandy soil slope is
easier to be destroyed by seepage flow than a clayey soil slope, so more attention should
be paid to the prevention of seepage flow from destroying a sandy soil slope. Actually,
Hong Kong is famous for slope failures and there are many sandy slopes failures (soil is
Chapter 8: Stability Analysis of Slope with Water Flow by Strength Reduction Method

- 179 -
completely decomposed granite) during raining time each year. On the other hand, there
is much less slope failures in those slopes with finer grain size (completely decomposed
volcanic) in Hong Kong (GEO 1996) based on thirty years observations in Hong Kong.
The sandy and finer grain soils in Hong Kong are derived from granitic and volcanic
rock with similar chemical composition, and the main differences between the two types
of soil are the grain size and the cohesive strength. It is also found that the location of
the slip surface for sandy soil is greatly influenced by the water flow from various
parametric studies, and the failure surface becomes shallower and closer to the slope toe
under the influence of water seepage flow. These results are also similar to the
observations of the slopes failures in Hong Kong over the last thirty years (GEO 1996),
where many slopes failures are initiated from toe failures under heavy rain.

In the above analysis, the pore water pressure is generated by a seepage flow analysis
which is a reasonable way to obtain the pore pressure distribution. It is however
possible to define the pore water pressure by a water table in SRM similar to that in the
LEM analysis. In order to investigate the difference between the two approaches,
another model is developed in which the pore pressure is generated by the water table
and the pore pressure distribution is shown in Figure 8.6 (the water table location is the
free-surface obtained from Figure 8.1). The factors of safety for the two cases are
compared in Table 8.1. The FOS for the case where the pore pressure is generated by
the water table is smaller than the case generated by the seepage flow analysis, as the
pore water pressure calculated by the water table (free-surface) is larger. It means that
the use of the water table (or piezometric line) is a conservative method of analysis, and
the difference is very small for a clayey soil slope but is much larger for a sandy soil.


Chapter 8: Stability Analysis of Slope with Water Flow by Strength Reduction Method

- 180 -
8.3 Stability analysis for a simple slope with irregular pore pressure
In this section, two more models are analyzed for the 6m height slope with 45 slope
angle discussed in the previous section, but the pore pressure distribution is irregular
due to some blocking effects in the slope. In the first model, the soil on the left side of
the slope crest and on the left side of slope toe is assigned as impermeable zone. The
slip surface and the pore pressure distribution are shown in Figure 8.7. In the second
model, an 8 m height thin impermeable wall near the slope crest is applied. The slip
surface and the pore pressure distribution are shown in Figure 8.8. In the second model,
the FOS (1.30) is much larger than the case with no soil blocking in Figure 8.3 (0.96)
and the case in Figure 8.7 (1.07). With the blockage from the soil wall, the seepage path
gets much longer which will greatly reduce the pore water pressure, so the FOS gets
much larger. This result also shows that the installation of retaining wall to increase the
seepage length is a good method to prevent slope failure caused by seepage flow.

8.4 Stability analysis for soil nailing slope with seepage flow
In this section, a 6m height soil nailed slope with water flow is analyzed and the slope
angle is 45. Nails are installed on 1.5m centers horizontally and vertically. The nail
length is 8m and the inclination angle is zero. The cohesion of the soil is 2 kPa and the
friction angle is 35, and the slope model is shown in Figure 8.9. In Hong Kong where
the density of soil nail is high (spacing around 1m to 1.5m) and the diameter and length
of the soil nails are large (32mm or 40mm diameter bar is commonly used while a
length of 10m to 20m is also common), many engineers have questioned that the
densely populated soil nails may affect the seepage pattern. Towards this problem, two
models are developed. In the first model, the pore pressure is determined without the
presence of the nail. In the second model, the soil nails together with the grout are
Chapter 8: Stability Analysis of Slope with Water Flow by Strength Reduction Method

- 181 -
treated as impermeable zone and the pore pressure distribution for this case is shown in
Figure 8.10. For these two models, the soil is given a relatively large permeability
(0.01cm/s). In the first model, the nail is given the same permeability as the soil, while
in the second model, the nail is given a very small permeability (10
-8
cm/s). The water
boundary conditions are: at the right hand side, the water height is 10m; at the slope toe
side, the water height is 4m (Figure 8.10b). It is noticed that the pore pressure
distribution considering the blockage effect from soil nails is actually close to that for
the first case at the section where there is no soil nail. The FOS by these two models is
also the same (both are 0.95 and the slip surface as shown in Figure 8.11 apply for both
cases). This means that the blockage effect of the soil nail can be ignored in a soil nailed
slope analysis as the nail diameter is small as compared with the nail spacing, even
when large diameter soil nails as used in Hong Kong are used in the analysis.

For this soil nailed slope, the FOS is 1.71 with no water flow. If no nail is included in
this slope, the FOS with and without water flow is 0.72 and 1.07 respectively. It can be
seen that the factor of the safety for soil nailed slope is more influenced by the water
flow, as the percentage decrease of the FOS is much larger when compared with a slope
without nail. In this analysis, the nail is simulated by cable element and its pullout
strength is related to the confining pressure. When there is water, the confining pressure
around the nail is decreased due to the pore water pressure, so water has two effects on a
soil nailed slope. Firstly, the FOS is decreased by the seepage force, and secondly, the
FOS is lowered by the lowering of the nail pullout strength which is related to the
effective confining pressure. To remove the influence of the confining pressure on the
nail pullout strength, the second model is re-considered in which the nail has a constant
pullout strength which is not affected by the effective overburden stress (only grout
cohesive strength is given and grout friction is zero) and the results are shown in Table
Chapter 8: Stability Analysis of Slope with Water Flow by Strength Reduction Method

- 182 -
8.2. The decrease of the FOS with the presence of water is now smaller when compared
with the previous model. It is clear that the relation of the soil nail pullout strength and
the effective overburden stress is a very critical factor in the stability of a soil nailed
slope.

8.5 Stability analysis for piled slope with seepage flow
In this section, a piled slope with water flow is analyzed. The slope geometry and the
properties of soil and pile are the same as the first example analyzed in chapter 7
(Figure 7.1). Only the case of pile spacing equal to 3D is considered. Since the solution
time of SRM analysis for slope with water flow is much longer than for no water slope,
the model with a relatively coarse mesh is developed to reduce solution time, and the
mesh plot is shown in Figure 8.12. When the slope is not reinforced with pile, the factor
of safety by the SRM is 0.85 and the slip surface is shown in Figure 8.13.

Two models are developed to investigate the influence of seepage flow on the failure
mechanism for pile reinforced slope. The FOS and slip surface obtained by the models
with and without water blocking effect are the same for the two cases (FOS is 1.29 and
the results are shown in Figures 8.14 and 8.15). It is clear that the blocking effect of the
pile to the seepage can also be ignored which is similar to the case for soil nails.

If no water is involved in this model, it is found the optimal pile position is about at the
middle of the slope which is shown in chapter 7. The slip surface is practically divided
into two parts, and obvious shear strain is mobilized in both the lower and upper parts.
When there is water seepage, the slip surface is mainly located at the lower part of the
slope (Figure 8.15) which is different from the case without seepage, which means that
the upper part of the slope is safer than the lower part. This phenomenon clearly arises
Chapter 8: Stability Analysis of Slope with Water Flow by Strength Reduction Method

- 183 -
from the effect of the seepage force. Without the seepage force, there is only a minor
interaction between the upper and lower parts of the failure mass. The optimized pile
position under seepage moves towards the slope toe instead of locating at about the
middle of the slope, and this is different from that without seepage. It is found that the
optimal pile location for this case is 2.0m towards the slope toe as measured from the
middle of the slope (Figure 8.16). The effect of the water seepage is hence important in
controlling the failure mechanism of a pile reinforced slope.


8.6 Stability analysis for locally loaded slope with seepage flow
In this section, a 6m height slope with 45 slope angle under rectangular shape vertical
loading is analyzed. The width and length of the loading are 2m and 4m respectively,
while the edge of the loading is 1m away from the crest of the slope. The cohesion of
the soil is 20kPa and the friction angle is 20. The length of the computer model is 20m.
When the loading q is 100kPa, the results of analysis are shown in Figure 8.17. This
slope model with no water has been analyzed in chapter 4, and the result is shown in
Figure 4.16b. It can be seen from this model that the failure mechanisms for the slope
with water and without water are different. For the slope with a two-dimensional
seepage flow, the slip surface is basically two-dimensional. On the other hand, for the
slope without water, a nearly three-dimensional slip surface is mobilized around the
local loading. In chapter 4, the failure mechanism of the locally loaded slope with no
water has been investigated. When the loading is small, the slip surface is still basically
two-dimensional until the loading becomes sufficiently large to mobilize a three-
dimensional slip surface. For this model, the failure mechanism with and without
seepage are different even though the applied loading is the same because the seepage
force is included in the analysis, and the ability to mobilize a three-dimensional slip
surface for the local loading is weakened by the two-dimensional seepage flow.
Chapter 8: Stability Analysis of Slope with Water Flow by Strength Reduction Method

- 184 -
8.7 Conclusion
In this chapter, the strength reduction method is employed for slope stability analysis
with water flow. The pore water pressure is generated by seepage flow analysis
according to the boundary conditions. With seepage flow, the FOS is usually much
smaller than the corresponding case with no water. For the sandy soil slope, the
decrease of the FOS by seepage is usually larger than that for the clayey soil slope. This
means that sandy soil slope is easier to be destroyed by seepage flow than clayey soil
slope which agrees well with the observations in Hong Kong over the last thirty years.
The reduction of seepage flow in sandy soil slope hence deserves more attention.
Furthermore, the location of the slip surface for sandy soil slope is more sensitive to
seepage, and it becomes shallower and closer to the slope toe under the influence of
seepage flow. If the pore pressure is generated by the use of piezometric line, usually
the FOS will be smaller than the case where the pore pressure is generated by seepage
flow analysis. It means that the water table (or piezometric line) option is a conservative
method for analysis. For clayey soil, the difference between the two ways in defining
the pore water pressure is usually small, but for sandy soil, the difference is much
larger.

It has been demonstrated that the use of a retaining wall to increase the seepage path is a
very effective solution to enhance slope stability under seepage flow. On the other hand,
the results of the seepage analysis are virtually independent of the soil nails or
reinforcing piles (under practical spacing) so that engineers need not consider the soil
nail/pile in the seepage analysis. The water flow has two effects on the stability of a soil
nailed slope. Firstly, the FOS is decreased by the seepage force. Secondly, the FOS is
decreased by the reduction of the nail pullout strength as the confining pressure around
Chapter 8: Stability Analysis of Slope with Water Flow by Strength Reduction Method

- 185 -
the nail is decreased by the pore water pressure. Water is hence a major factor in
controlling the stability of slope.

For locally loaded slope with water flow and pile reinforced slope, the failure
mechanism can be strongly influenced by the seepage force. To get a realistic failure
mechanism, the pore pressure must be carefully considered in the analysis.




















Chapter 8: Stability Analysis of Slope with Water Flow by Strength Reduction Method

- 186 -

Table 8.1 Factor of safety for different situations by SRM
Cohesion and
Friction angle
1kPa,
45
2kPa,
45
5kPa,
35
10kPa,
25
FOS with water ( pore pressure
generated by seepage analysis)
0.82 0.96 0.98 1.07
FOS with water ( pore pressure
generated by water table)
0.66 0.82 0.92 1.03
FOS for no water 1.3 1.44 1.35 1.37


Table 8.2 Factor of safety of nailed slope by SRM
Case No water With water flow
With no nail 1.07 0.72
Nail pull out strength related to confining pressure 1.71 0.95
Nail with constant pull out strength 5.6kN/m 1.61 1.01
















Chapter 8: Stability Analysis of Slope with Water Flow by Strength Reduction Method

- 187 -


10m
4m

Figure 8.1 Pore water pressure and flow vector of a simple slope from a free-surface
seepage analysis


(a) with water flow, FOS=0.82 (b) no water, FOS=1.30
Figure 8.2 Slip surface for slope with cohesion 1kPa and friction angle 45


(a) with water flow, FOS=0.96 (b) no water, FOS=1.44
Figure 8.3 Slip surface for slope with cohesion 2kPa and friction angle 45



Chapter 8: Stability Analysis of Slope with Water Flow by Strength Reduction Method

- 188 -


(a) with water flow, FOS=0.98 (b) no water, FOS=1.35
Figure 8.4 Slip surface for slope with cohesion 5kPa and friction angle 35


(a) with water flow, FOS=1.07 (b) no water, FOS=1.37
Figure 8.5 Slip surface for slope with cohesion 10kPa and friction angle 25


Figure 8.6 Water pressure by water table (piezometric line) assuming hydrostatic
condition


Chapter 8: Stability Analysis of Slope with Water Flow by Strength Reduction Method

- 189 -


(a) slip surface, FOS=1.07 (b) pore water pressure
Figure 8.7 Slip surface and pore pressure with water block at upper and bottom left for
slope with cohesion 2kPa and friction angle 45


(a) slip surface, FOS=1.31 (b) pore water pressure
Figure 8.8 Slip surface and pore pressure with water block wall for slope with cohesion
2kPa and friction angle 45


Figure 8.9 Model plot of the soil nailing slope

Chapter 8: Stability Analysis of Slope with Water Flow by Strength Reduction Method

- 190 -


10m
4m

(a) cutting plane through the nail (b) cutting plane not including nail
Figure 8.10 Pore pressure distribution of the soil nailing slope with water blocking
effect


Figure 8.11 Slip surface for the nailed slope with water flow (FOS=0.95)


Figure 8.12 Model plot of the piled slope

Chapter 8: Stability Analysis of Slope with Water Flow by Strength Reduction Method

- 191 -


(a) pore pressure (b) slip surface
Figure 8.13 Pore pressure and slip surface of the slope without pile (FOS=0.85)



(a) cutting plane through the pile center (b) cutting plane through soil midway
Figure 8.14 Pore pressure distribution of the piled slope with water blocking effect



(a) back view (b) front view
Figure 8.15 Slip surface for the piled slope with water flow (FOS=1.29)




Chapter 8: Stability Analysis of Slope with Water Flow by Strength Reduction Method

- 192 -


(a) model plot (b) slip surface
Figure 8.16 Slip surface for the slope with pile installed at 2.0m towards the slope toe as
measured from the middle of slope (FOS=1.34)


(a) pore pressure distribution (b) slip surface, FOS=1.29
Figure 8.17 Pore water pressure and slip surface for the locally loaded slope with water
Chapter 9: Conclusions and Recommendations

- 193 -

CHAPTER 9: CONCLUSIONS AND RECOMMENDATIONS


9.1 Conclusions
This study has attempted to conduct a comprehensive analysis of three-dimensional
slopes. The 3D effects include irregular slope geometry, soil property distribution and
boundary conditions as well as the effects of soil nail/pile and water flow. Both limit
equilibrium and strength reduction methods have been used in this study, and an
extensive comparison between these two methods are also conducted. The main
conclusions of the present study are summarized in five aspects which are described as
follows.

9.1.1 Comments on limit equilibrium and strength reduction methods
The main advantage of strength reduction method is that it can automatically locate the
critical slip surface, and the failure modes can be detected by the shear strain
distribution. Another advantage of SRM is that it can easily simulate more complex
conditions with inclusions. Concerning the piled slope, the SRM can be a better
alternative to the LEM, as the LEM is still not very mature for piled slope analysis. This
advantage may be more pronounced when different reinforcements are included in the
same slope, and a complicated water flow conditions is involved in the analysis.

The major limitations of SRM are: it is sensitive to the nonlinear solution algorithm for
the case of a soft band especially for frictional material, and it is incapable of
determining other failure surfaces which may be only slightly less critical than the SRM
solution but still require treatment for good engineering practice. Another issue is that
the SRM can be sensitive to the convergence criterion, and if it is necessary, the result
Chapter 9: Conclusions and Recommendations

- 194 -
should be checked by the criterion of sudden change of displacement which is found to
be a more reliable criterion to determine the failure state but is not the default option in
some commercial program (but require human judgment instead of automatic analysis).
The long solution time for SRM has been a major problem in the past, but this problem
becomes moderate now with the continuing improvement of computational power.

The main shortcoming of the LEM is that it includes many assumptions to make the
solution statically determinate. It is also not as powerful as SRM to simulate interaction
between reinforcement and soil, and to analyze some very complex situations. The
primary benefit of the LEM is the short solution time, and it is still more suitable and
generally accepted for routine design and analysis. The LEM is also flexible in finding
different local minima factor of safety and the corresponding slip surface. Through the
present study, it can be concluded that both the LEM and SRM have their own merits
and demerits, and the final choice will heavily depends on the engineering judgment of
the results.

9.1.2 Main findings from analysis of non-reinforced slopes with obvious 3D effects
The results from the 3D LEM and SRM agree reasonably well for normal situations,
and the factors of safety obtained by the SRM are slightly higher than those from the
LEM, which is similar to the corresponding 2D analysis. Such results also agree well
with previous studies by other researchers.

For an infinite slope with local loading, whether the critical failure surface is 2D or 3D
depends on the magnitude of the external load. It is also interesting to find that the
suitable domain size for analysis is largely controlled by the failure mechanism instead
of the length of the external load, as the failure mechanism may be greatly affected by
Chapter 9: Conclusions and Recommendations

- 195 -
the boundary effect. For a proper SRM or LEM analysis, the engineers may either adopt
a large domain size at the expense of computer time, or pursue a trial and error analysis
as discussed in previous chapters to determine the proper domain size.

For concave slope, the effect of curvature is distinct and the FOS increases with the
decrease of radius of curvature. For convex slope with no local loading, the curvature
effect can be ignored, but for the steep convex slope with local loading, the curvature
effect is apparent.

9.1.3 Main findings from analysis of soil nailing slope
For soil nailed slope, it is also found that there are not major differences in the factor of
safety and slip surfaces from the strength reduction method and the limit equilibrium
method in general. Appreciable difference between the SRM and LEM will however be
found if the nail load is controlled by the overburden stress.

It is demonstrated that a nail head is important in determining the failure mode and the
factor of safety of a nailed slope, while the effect of the nail elastic modulus is more
noticeable only when the slope is very steep.

The optimum layout of the soil nail is found to be longer at the bottom of the slope and
shorter at the top which is contrary to some of the engineers guideline for design of soil
nail during top-down construction. In general, longer soil nail should be placed at the
location where failure first initiates.

Chapter 9: Conclusions and Recommendations

- 196 -
It is found that the mobilized bending moment in soil nail is small even when the nail
inclined angle is large. In this respect, the conventional method of soil design appears to
be adequate and is slightly on the conservative side.

From the tension force distribution along the soil nail, it is found that the tension force
distribution is influenced by the state of the slope (service state, limit state) and the
failure modes (external failure, internal failure). In general, the line of the maximum
tension may not correspond to the critical slip surface as commonly believed, except
when the failure mode is an internal tensile failure. This result is contrary to the
common belief by the engineers, but the detailed study in chapter 5 has demonstrated
that this is actually a better representation of the nail load distribution.

For locally loaded slopes, the nail force of the top row is usually more mobilized, but
for non-locally loaded slopes, the nail force of the bottom row is usually more
mobilized at the limit state.

For the reinforcement design of intersected slope, less reinforcement can be used for the
concave intersection part as the effect of curvature can improve the slope stability. For
convex intersection slope, if the slope angle is not high, the reinforcement design can be
the same as no curvature slopes. If the slope is steep, more reinforcement should be
used for convex intersection slope.

9.1.4 Main findings from analysis of piled slope
For slope reinforced with one row of piles, the slip surface is found to be divided into
two parts when the pile spacing is very small, and these two parts gradually get
connected with the increase of pile spacing until a clear single slip surface is formed
Chapter 9: Conclusions and Recommendations

- 197 -
when the pile spacing is large enough. The slip surface of the piled slope is usually
shallower than the corresponding slope with no pile, and this result is totally different
from the previous results based on the maximum point of shear force where a very deep
slip surface for piled slope is usually obtained.

The optimal pile position for slope reinforced with one row of piles is found to locate
between the middle of the slope and the middle of the slip surface for an unreinforced
slope.

An optimal design procedure for the pile spacing has also been suggested in the present
study based on the upper bound and lower bound of the factor of safety, and the
procedure will help to reduce the time required for a piled slope three-dimensional
analysis.

9.1.5 Main findings from analysis of slopes with water flow
From the analysis, it demonstrates that the factor of safety for sandy soil slope is more
influenced by seepage flow. Thus sandy soil slope deserves more attention in prevention
of destroy from water flow. This phenomenon is also in compliance with the
observation in Hong Kong where less slope failures are found for slopes with
completely decomposed volcanic which has higher cohesive strength than the slopes
with completely decomposed granite.

If the pore pressure is generated by the use of piezometric line, the FOS will be smaller
than that generated by seepage flow analysis. The difference is small for clayey soil
slope but is much larger for sandy soil slope. This result has demonstrated that the
classical method in treating the water table is conservative for analysis and design.
Chapter 9: Conclusions and Recommendations

- 198 -
The analysis also shows that the installation of retaining wall to increase the length of
the seepage path is a good method to prevent slope failure induced by seepage flow.

The water flow has two effects on the soil nailing slope stability. Firstly, the FOS is
decreased by the seepage force and secondly, the FOS is decreased by the weakening of
the nail pullout strength as the effective confining pressure around the nail is decreased
by the pore water pressure.

For locally loaded slope with water flow, the failure mechanism is influenced by the
seepage force. The three-dimensional slip surface is more difficult to be mobilized when
a two-dimensional seepage force is included.

The present study also shows that the effect of pile and densely populated soil nail on
the seepage flow can be neglected for practical purpose. This result is important to
many engineers in Hong Kong, as this question has been raised among the engineers in
Hong Kong due to the densely populated soil nails used locally.

9.2 Recommendations and suggestions
For problems with a soft band or major differences in the soil parameters between soil
layers, the numerical model and results should be checked carefully. Since both the
SRM and LEM have their specific advantages and limitations, the results come from the
complex slope models should be cross checked by these two methods. The final results
have to be assessed carefully using engineering judgment.

It is found that the non-convergence criterion in commercial SRM programs should be
assessed carefully, and engineers should not rely completely on the default setting in the
Chapter 9: Conclusions and Recommendations

- 199 -
computer programs. It has also been found that a sudden change of the displacement
may sometimes be a better criterion than the default setting in the computer programs
for some cases. Therefore, more research to establish a more robust criterion for the
ultimate limit state under general condition is required. A proper and successful
combination of the above two criteria may be a prospective way.

For piled slope analysis, only one row of piles is considered and infinite pile length
assumption is used in this study. Further research should be conducted on the slope with
multi rows of piles, and the influence of the pile length on the slope failure mode should
also be investigated.

In the present study, only one kind of reinforcement (soil nail or pile) is included in an
individual slope. Further study should be carried out on the stability and failure modes
of the slope with the presence of different types of reinforcement.

Many 3D effects and failure modes are revealed by numerical simulation in the present
study, and some of them are verified by limit equilibrium method or testing results
obtained in the literature. Further study is still suggested by conducting full scale or
reduced scale model tests to confirm some of the results from the numerical simulation.
References

- 200 -

REFERENCES
Abramson, L. W., Lee, T. S., Sharma, S., and Boyce, G. M. (2002). Slope Stability and
Stabilization Methods, 2nd edition, J ohn Wiley, USA
Anderheggen, E. and Knopfel, H. (1972). Finite element limit analysis using linear
programming. International J ournal of Solids and Structures, 8, 1413-1431
ASCE. (1997). Committee report Soil nailed retaining structures. Ground
Improvement, Ground Reinforcement, Ground Treatment: developments 1987-
1997, ASCE Geotechnical Special Publication No. 69, edited by Vernon R.
Schaefer, Section 2.8, 201-232
Ausilio, E., Conte, E., and Dente, G. (2001). Stability analysis of slopes reinforced with
piles. Computers and Geotechnics, 28(8), 591-611
Azzouz, A. S. and Baligh, M. M. (1983). Loaded areas on cohesive slopes. J ournal of
Geotechnical Engineering, ASCE, 109(5), 724-729
Babu, G. L. S., Murthy, B. R. S., and Srinivas, A. (2002). Analysis of construction
factors influencing the behaviour of soil-nailed earth retaining walls. Ground
Improvement, 6(3), 137-143
Baker, R. and Garber, M. (1978). Theoretical analysis of the stability of slopes.
Geotechnique, 28(4), 395-411
Baker, R., and Leshchinsky, D. (1987). Stability analysis of conical heaps. Soils and
Foundations, 27(4), 99-110
Baker, R. and Leshchinsky, D. (2001). Spatial distribution of safety factors. J ournal of
Geotechnical and Geoenvironmental Engineering, 127(2), 135-145
Baligh, M. M. and Azzouz, A. S. (1975). End effects on stability of cohesive slopes.
J ournal of the Geotechnical Engineering Division, ASCE, 101(GT11), 1105-1118
References

- 201 -
Bell, J . M. (1968). General slope stability analysis. J ournal of the Soil Mechanics and
Foundations Division, ASCE, 94(SM6), 1253-1270
Bishop, A. W. (1955). The use of the slip circle in stability analysis of slopes.
Geotehnique, 5(1), 7-17
Bottero, A., Negre, R., Pastor, J ., and Turgeman, S. (1980). Finite element method and
limit analysis theory for soil mechanics problems. Computer Methods in Applied
Mechanics and Engineering, 22, 131-149
Briaud, J . L. and Lim, Y. (1997). Soil-nailed wall under piled bridge abutment:
simulation and guidelines. J ournal of Geotechnical and Geoenvironmental
Engineering, 123(11), 1043-1050
Brinkgreve, R. B. J . and Bakker, H. L. (1991). Non-linear finite element analysis of
safety factors. Proceedings of the seventh international conference on computer
methods and advances in geomechanics, Cairns, 6-10 May 1991, 1117-1122
Bruce, D. A. and J ewell, R. A. (1986). Soil nailing: Application and Practice-Part 1.
Ground Engineering, 19(8), 10-15
Byrne, R. J ., Cotton, D., Porterfield, J ., Wolschlag, C., and Ueblacker, G. (1998).
Manual for Design & Construction of Soil Nail Walls. FHWA-SA-96-069R,
Federal Highway Administration (FHWA), Washington DC, USA.
Cai, F., Ugai, K., Wakai, A., and Li, Q. (1999). Effects of horizontal drains on slope
stability under rainfall by three-dimensional finite elements analysis. Computers
and Geotechnics, 23, 255-275
Cai, F. and Ugai, K. (2000). Numerical analysis of the stability of a slope reinforced
with piles. Soils and Foundations, 40(1), 73-84
Cai, F. and Ugai, K. (2003). Response of flexible piles under laterally linear movement
of the sliding layer in landslides. Canadian Geotechnical J ournal, 40(1), 46-53
References

- 202 -
Cai, F. and Ugai, K. (2004). Numerical analysis of rainfall effects on slope stability.
International J ournal of Geomechanics, 4(2), 69-78
Cartier, G. and Gigan, J . P. (1983). Experiments and observations on soil nailing
structures. Proc. Enropean Conf. Soil Mechanics and Foundation Engineering,
Helsinki, 473-476
Cavounidis, S. (1987). On the ratio of factors of safety in slope stability analyses.
Geotechnique, 37 (2), 207- 210
Chai, X. J . and Hayashi, S. (2005). Effect of constrained dilatancy on pull-out resistance
of nails in sandy clay. Ground Improvement, 9(3), 127-135
Chang, M. (2002). A 3D slope stability analysis method assuming parallel lines of
intersection and differential straining of block contacts. Canadian Geotechnical
J ournal, 39, 799-811
Chen, C. Y. and Martin, G. R. (2002). Soil-structure interaction for landslide stabilizing
piles. Computers and Geotechnics, 29(5), 363-386
Chen, L. T. and Poulos, H. G. (1997). Piles subjected to lateral soil movements. J ournal
of Geotechnical and Geoenvironmental Engineering, 123(9), 802-811
Chen, Z., Wang, X., Haberfield, C., Yin, J . H., and Wang, Y. (2001a). A three-
dimensional slope stability analysis method using the upper bound theorem, Part 1:
theory and methods. International J ournal of Rock Mechanics and Mining Sciences,
38, 369-378
Chen, Z., Wang, J ., Wang, Y., Yin, J . H., and Haberfield, C. (2001b). A three-
dimensional slope stability analysis method using the upper bound theorem, Part 2:
numerical approaches, applications and extensions. International J ournal of Rock
Mechanics and Mining Sciences, 38, 379-397
References

- 203 -
Chen, Z. Y., Mi, H. L., Zhang, F. M., and Wang, X. G. (2003a). A simplified method
for 3D slope stability analysis. Canadian Geotechnical J ournal, 40, 675-683
Chen, J . (2004). Slope stability analysis using rigid elements. PhD Thesis, The Hong
Kong Polytechnic University
Chen, J ., Yin, J . H., and Lee, C. F. (2003b). Upper bound limit analysis of slope
stability using rigid finite elements and nonlinear programming. Canadian
Geotechnical J ournal, 40, 742-752
Chen, J ., Yin, J . H., and Lee, C. F. (2004). Rigid finite element method for upper bound
limit analysis of soil slopes subjected to pore water pressure. J ournal of Engineering
Mechanics, 130(8), 886-893
Chen, J ., Yin, J . H., and Lee, C. F. (2005a). The use of an SQP algorithm in slope
stability analysis. Communications in Numerical Methods in Engineering, 21, 23-37
Chen, J ., Yin, J . H., and Lee, C. F. (2005b). A three-dimensional upper-bound approach
to slope stability analysis based on RFEM. Geotechnique, 55(7), 549-556
Chen, R. H. and Chameau, J . L. (1982). Three-dimensional limit equilibrium analysis of
slopes. Geotechnique, 32(1), 31-40
Chen, W. F. (1975). Limit Analysis and Soil Plasticity. Elsevier Scientific Publishing
Co., New York
Chen, W. F. and Giger, M. W. (1971). Limit analysis of stability of slopes. J ournal of
the Soil Mechanics and Foundations Division, 97(1), 19-26
Cheng, Y. M., Liu, H. T., Wei, W. B., and Au, S. K. (2005). Location of critical three-
dimensional non-spherical failure surface by Nurbs Functions and Ellipsoid with
applications to highway slopes. Computers and Geotechnics, 32, 387-399
Cheng, Y. M. (2003). Locations of critical failure surface and some further studies on
slope stability analysis. Computers and Geotechnics, 30, 255-267
References

- 204 -
Cheng, Y. M. and Zhu, L. J . (2004). Unified formulation for two dimensional slope
stability analysis and limitations in factor of safety determination. Soils and
Foundations, 44(6), 121-127
Cheng, Y. M. and Yip, C. J . (2007). Three-dimensional asymmetrical slope stability
analysis extension of Bishops, J anbus, and Morgenstern-Prices techniques.
J ournal of Geotechnical and Geoenvironmental engineering, ASCE, 133(12),
1544-1555
Cheng, Y. M., Li, L. and Chi, S. C. (2007). Performance studies on six heuristic global
optimization methods in the location of critical slip surface. Computers and
Geotechnics, 34(6), 462-484
Cheng, Y. M. (2007). Global optimization analysis of slope stability by simulated
annealing with dynamic bounds and Dirac function, Engineering Optimization,
39(1), 17-32
Cheng, Y. M., Zhao, Z. H., and Wang, J . A. (2008). Realization of Pan J iazhengs
extremum principle with optimization methods. Chinese J ournal of Rock
Mechanics and Engineering, 27(4), 782-788 (in Chinese)
Cheuk, C. Y., Ng, C. W. W., and Sun, H. W. (2005). Numerical experiments of soil
nails in loose fill slopes subjected to rainfall infiltration effects. Computers and
Geotechnics, 32(4), 290-303
Cho, S. E. and Lee, S. R. (2001). Instability of unsaturated soil slope due to infiltration.
Computers and Geotechnics, 28, 185-208
Choukeir, M. H. (1996). Seismic analysis of reinforced earth and soil-nailed structures.
Ph.D Thesis, New York Polytechnic University
Chow, Y. K. (1996). Analysis of piles used for slope stabilization. International J ournal
for Numerical and Analytical Methods in Geomechanics, 20, 635-646
References

- 205 -
Chu, L. M. (2003). Study on the interface shear strength of soil nailing in completely
decomposed granite (CDG) soil. M.Phil thesis, Sept. 2003, The Hong Kong
Polytechnic University
Chu, L. M. and Yin, J . H. (2005). Comparison of interface shear strength of soil nails
measured by both direct shear box tests and pullout tests. J ournal of Geotechnical
and Geoenvironmental Engineering, 131(9), 1097-1107
Clouterre. (1993). Recommendations Clouterre 1991. US Department of Transportation,
Federal Highway Administration (English translation, Report on the French National
Project Clouterre, Rep. No. FHWA-SA-93-026, Washington, D.C.)
DAppolonia, E., Alperstein, R., and DAppolonia, D. J . (1967). Behavior of a colluvial
slope. J ournal of the Soil Mechanics and Foundation Division, ASCE, 93(SM4),
447-473
Dawson, E. M., Roth, W. H., and Drescher, A. (1999). Slope stability analysis by
strength reduction. Geotechnique, 49(6), 835-840
Dawson, E., Motamed, F., Nesarajah, S., and Roth, W. (2000). Geotechnical stability
analysis by strength reduction. Slope stability 2000: proceedings of sessions of
Geo-Denver 2000, August 5-8, 2000, Denver, Colorado, 99-113
De Beer, E. E. and Wallays, M. (1970). Stabilization of a slope in schist by means of
bored piles reinforced with steel beams. Proc. 2
nd
Int. Congress Rock Mech., Vol.
3, 361-369
Deng, J . H., Tham, L. G., Lee, C. F., and Yang, Z. Y. (2007). Three-dimensional
stability evaluation of a preexisting landslide with multiple sliding directions by the
strength-reduction technique. Canadian Geotechnical J ournal, 44, 343-354
Donald, I. and Chen, Z. Y. (1997). Slope stability analysis by the upper bound approach:
fundamentals and methods. Canadian Geotechnical J ournal, 34, 853-862
References

- 206 -
Donald, I. B. and Giam, S. K. (1988). Application of the nodal displacement method to
slope stability analysis. Proceedings of the 5th Australia-New Zealand conference
on geomechanics, Sydney, Australia, 456-460
Drucker, D. C. and Prager, W. (1952). Soil mechanics and plastic analysis or limit
design. Quarterly of Applied Mathematics, 10, 157-165
Duncan, J . M. (1996). State of the art: limit equilibrium and finite-element analysis of
slopes. J ournal of Geotechnical Engineering, 122(7), 577-596
Duncan, J . M. and Wright, S. G. (2005). Soil Strength and Slope Stability. J ohn Wiley
& Sons, Inc.
Elias, V. and Christopher, B. R. (1997). Mechanically Stabilized Earth Walls and
Reinforced Soil Slopes Design and Construction Guidelines. FHWA DM 82
Manual, Report No. FHWA A-SA-96-071, USA
Elias, V., Christopher, B. R., and Berg, R. R. (2001). Mechanically Stabilized Earth
Walls and Reinforced Soil Slopes Design and Construction Guidelines. Federal
Highway Administration (FHWA), FHWA-NHI-00-043, USA
Farzaneh, O. and Askari, F. (2003). Three-dimensional analysis of nonhomogeneous
slopes. J ournal of Geotechnical and Geoenvironmental Engineering, 129(2), 137-
145
Fredlund, D. G. and Krahn, J . (1984). Analytical methods for slope analysis.
International Symposium on Landslides, 229-250
Fredlund, D. G., Scoular, R. E. G., and Zakerzadeh, N. (1999). Using a finite element
stress analysis to compute the factor of safety. Proceedings of the 52
nd
Canadian
Geotechnical Conference, Regina, Saskatchewan, 73-80
Fukuoka, M. (1977). The effects of horizontal loads on piles due to landslides. Proc.
10
th
Spec. Session, 9
th
Int. Conf. Soil Mech. and Found. Eng., Tokyo, 27-42
References

- 207 -
Gasmo, J . M., Rahardjo, H., and Leong, E. C. (2000). Infiltration effects on stability of
a residual soil slope. Computers and Geotechnics, 26, 145-165
Gassler, G. and Gudehus, G. (1981). Soil nailing Some aspects of a new technique.
Proceedings of the Tenth International Conference on Soil Mechanics and
Foundation Engineering, Sweden, Vol. 3, 665-670
Gassler, G. (1988). Soil-nailing - theoretical basis and practical design. Pro. Int.
Geotechnical Symp. on Theory and Practice of Earth Reinforcement, Fukuoka
Kyushu, 283288
Gassler, G. (1992). Full scale test on a nailed wall in consolidated clay. Proc. Int. Symp.
on Earth Reinforcement Practice, Fukuoka, Kyushu, J apan, Vol. 1, 475-480
Gassler, G. (1993). The first two field tests in the history of soil nailing on nailed walls
pushed to failure. Soil reinforcement: full scale experiments of the 80s, Presses de
lecole Nationale des Ponts et Chaussees, CEEC, 7-34
Gens, A., Hutchinson, J . N., and Cavounidis, S. (1988). Three-dimensional analysis of
slides in cohesive soils. Geotechnique, 38(1), 1-23
GEO. (1996). GEO Report No. 52: Investigation of Some Major Slope Failures between
1992 and 1995. The Hong Kong Special Administrative Region Government
Giger, M. W. and Krizek, R. J . (1975). Stability analysis of vertical cut with variable
conner angle. Soils and Foundations, 15(2), 63-71
Giger, M. W. and Krizek, R. J . (1976). Stability of vertical corner cut with concentrated
surcharge load. J ournal of the Geotechnical Engineering Division, 102(1), 31-40
Greenwood, J . R. (1983). A simple approach to slope stability. Ground Engineering,
16(4), 45-48
Greenwood, J . R. (1985). Correspondence on stability of compacted rockfill slopes,
Charles, J . A. and Soares, M. M. (1984). Geotechnique, 35(2), 217-218
References

- 208 -
Griffiths, D. V. and Lane, P. A. (1999). Slope stability analysis by finite elements.
Geotechnique, 49(3), 387-403
Griffiths, D. V. and Marquez, R. M. (2007). Three-dimensional slope stability analysis
by elasto-plastic finite elements. Geotechnique, 57(6), 537-546
Gurocak, Z., Alemdag, S., and Zaman, M. M. (2008). Rock slope stability and
excavatability assessment of rocks at the Kapikaya dam site, Turkey. Engineering
Geology, 96, 17-27
Hassiotis, S., Chameau, J . L., and Gunaratne, M. (1997). Design method for
stabilization of slopes with piles. J ournal of Geotechnical and Geoenvironmental
Engineering, 123(4), 314-323
Hayashi, S., Ochiai, H., Otani, J ., and Umezaki, T. (1992). Function and evaluation of
steel bars in earth reinforcement. Proc. of the Int. Symp. on Earth Reinforcement,
Kyushu, J apan, Vol. 1, 481-486
Hong, Y. S., Wu, C. S., and Yang, S. H. (2003). Pullout resistance of single and double
nails in a model sandbox. Canadian Geotechnical J ournal, 40, 1039-1047
Hovland, H. J . (1977). Three-dimensional slope stability analysis method. J ournal of the
Geotechnical Engineering Division, ASCE, 103(GT9), 971-986
Huang, C. C. and Tsai, C. C. (2000). New method for 3D and asymmetrical slope
stability analysis. J ournal of Geotechnical and Geoenvironmental Engineering,
ASCE, 126(10), 917-927
Huang, C. C. and Tsai, C. C. (2002). General method for three-dimensional slope
stability analysis. J ournal of Geotechnical and Geoenvironmental Engineering,
ASCE, 128(10), 836-848
Hungr, O. (1987). An extension of Bishops simplified method of slope stability
analysis to three dimensions. Geotechnique, 37(1), 113-117
References

- 209 -
Itasca Consulting Group, Inc. (2006). FLAC 3D Version 3.1 Users Guide. Minneapolis,
Minnesota, USA
Ito, T. and Matsui, T. (1975). Methods to estimate lateral force acting on stabilizing
piles. Soils and Foundations, 15(4), 43-60
Ito, T., Matsui, T., and Hong, W. P. (1981). Design method for stabilizing piles against
landslide - one row of piles. Soils and Foundations, 21(1), 21-37
Ito, T., Matsui, T., and Hong, W. P. (1982). Extended design method for multi-row
stabilising piles against landslide. Soils and Foundations, 22(1), 1-13
Izbicki, R. J . (1981). Limit plasticity approach to slope stability problems. J ournal of the
Geotechnical Engineering Division, 107(2), 228-233
J anbu, N. (1973). Slope stability computation. Embankment Dam Engineering
Cassagrande Volume, J ohn Wiley
J ewell, R. A. (1990). Review of theoretical models of soil nailing. Proc. Int. Reinforced
Soil Conf., Glasgow, Scotland, 265-275
J ewell, R. A. and Pedley, M. J . (1992). Analysis for soil reinforcement with bending
stiffness. J ournal of Geotechnical Engineering Division, 118(10), 1505-1528
J iang, J . C. and Yamagami, T. (1999). Determination of the sliding direction in three-
dimensional slope stability analysis. Proc. 44
th
Symposium on Geotechnical
Engineering, 193-200
J ong, G. D. J . D. (1980). Application of the calculus of variations to the vertical cut off
in cohesive frictionless soil. Geotechnique, 30(1), 1-16
J unaideen, S. M., Tham, L. G., Law, K. T., Lee, C. F., and Yue, Z. Q. (2004).
Laboratory study of soil-nail interaction in loose, completely decomposed granite.
Canadian Geotechnical J ournal, 41, 274-286
J uran, I., Beech, J ., and De Laure, E. (1984). Experimental study of the behaviour of
nailed soil retaining structures on reduced scale models. Proceedings of
References

- 210 -
International Symposium on In Situ Soil and Rock Reinforcement, Paris, France,
309-314 (in French)
J uran, I., Shafiee, S., and Schlosser, F. (1985). Numerical study of nailed soil retaining
structures. Proceedings of the Eleventh International Conference on Soil Mechanics
and Foundation Engineering, San Franciso, Vol. 3, 1713-1716
J uran, I. and Elias, V. (1987). Soil nailed retaining structures: analysis of case histories,
Soil improvement A ten year update, ASCE Special Geotechnical Publication No.
12, Edited by J oseph P. Welsh, 232-244
J uran, I., Baudrand, G., Farrag, K., and Elias, V. (1990). Kinematical limit analysis for
design of soil-nailed structures. J ournal of Geotechnical Engineering, 116(1), 54-72
J uran, I., Baudrand, G., Farrag, K. and Elias, V. (1992). Kinematical limit analysis for
design of soil-nailed structures (discussion). J ournal of Geotechnical Engineering,
118, 1640-1647
Karal, K. (1977a). Application of energy method. J ournal of the Geotechnical
Engineering Division, 103(5), 381-397
Karal, K. (1977b). Energy method for soil stability analyses. J ournal of the
Geotechnical Engineering Division, 103(5), 431-445
Kim, D. S., J uran, I., Nasimov, R., and Drabkin, S. (1995). Model study on the failure
mechanism of soil-nailed structure under surcharge loading. Geotechnical Testing
J ournal, 18(4), 421-430
Kim, J . S., Park, C. L., Lee, S. D., and Lee, S. R. (1996). A large-scale experimental
study of soil-nailed structures. Proc. of the 3
rd
Int. Symp. on Earth Reinforcement,
Fukuoka, Kyushu, J apan, 775-780
Kim, J ., Salgado, R., and Yu, H. S. (1999). Limit analysis of soil slopes subjected to
pore- water pressures. J ournal of Geotechnical and Geoenvironmental Engineering,
125(1), 49-58
References

- 211 -
Kim, J ., Salgado, R., and Lee, J . (2002). Stability analysis of complex soil slopes using
limit analysis. J ournal of Geotechnical and Geoenvironmental engineering, ASCE,
128(7), 546-557
Kim, J ., J eong, S., Park, S., and Sharma, J . (2004). Influence of rainfall-induced wetting
on the stability of slopes in weathered soils. Engineering Geology, 75, 251-262
King, G. J . W. (1989). Revision of effective-stress method of slices. Geotechnique,
39(3), 497-502
Kitamura, T., Nagao, A., and Uehara, S. (1988). Model loading tests of reinforced
slopes with steel bars. Proc. Int. Geotechnical Symp. on Theory and Practice of
Earth Reinforcement, Fukuoka, J apan, 5-7 October, 311-316
Krahn, J . (2003). The 2001 R.M. Hardy Lecture: the limits of limit equilibrium analyses.
Canadian Geotechnical J ournal, 40(3), 643-660
Lam, L. and Fredlund, D. G. (1993). A general limit equilibrium model for three-
dimensional slope stability analysis. Canadian Geotechnical J ournal, 30, 905-919
Lazarte, C. A., Elias, V., Espinoza, R. D., and Sabatini, P. J . (2003). Geotechnical
Engineering Circular No. 7 - Soil Nail Walls. FHWA-IF-03-017, Federal Highway
Administration, USA
Lee, C. Y., Poulos, H. G., and Hull, T. S. (1991). Effect of seafloor instability on
offshore pile foundations. Canadian Geotechnical J ournal, 28(5), 729-737
Lee, C. Y., Hull, T. S., and Poulos, H. G. (1995). Simplified pile-slope stability
analysis. Computers and Geotechnics, 17(1), 1-16
Leshchinsky, D., Baker, R., and Silver, M. L. (1985). Three-dimensional analysis of
slope stability. International J ournal for Numerical and Analytical Methods in
Geomechanics, 9(2), 199-223
Leshchinsky, D. and Baker, R. (1986). Three-dimensional slope stability: end effects.
Soils and Foundations, 26(4), 98-110
References

- 212 -
Leshchinsky, D. and Huang, C. C. (1992a). Generalized slope stability analysis:
interpretation, modification, and comparison. J ournal of Geotechnical Engineering,
118(10), 1559-1576
Leshchinsky, D. and Huang, C. C. (1992b). Generalized three-dimensional slope
stability analysis. J ournal of Geotechnical Engineering, 118(11), 1748-1764
Li, J . (2003). Field study of a soil nailed loose fill slope. PhD thesis, J uly 2003, The
University of Hong Kong
Li, J ., Tham, L. G., J unaideen, S. M., Yue, Z. Q., and Lee, C. F. (2008). Loose fill slope
stabilization with soil nails: full-scale test. J ournal of Geotechnical and
Geoenvironmental Engineering, ASCE, 134(3), 277-288
Loehr, J . E., McCoy, B. F., and Wright, S. G. (2004). Quasi-three-dimensional slope
stability analysis method for general sliding bodies. J ournal of Geotechnical and
Geoenvironmental Engineering, 130(6), 551-560
Loukidis, D., Bandini, P., and Salgado, R. (2003). Stability of seismically loaded slopes
using limit analysis. Geotechnique, 53(5), 463-479
Lowe, J . and Karafiath, L. (1960). Stability of earth dams upon drawdown. Proceedings
of the 1th Pan-American on Soil Mechanics and Foundations Engineering,
Mexico City, Vol. 2, 537-552
Luo, S. Q., Tan, S. A., and Yong, K. Y. (2000). Pull-out resistance mechanism of a soil
nail reinforcement in dilative soils. Soils and Foundations, 40(1), 47-56
Luo, S. Q., Tan, S. A., Cheang, W., and Yong, K. Y. (2002). Elastoplastic analysis of
pull-out resistance of soil nails in dilatant soils. Ground Improvement, 6(4), 153-
161
Lyamin, A. V. (1999). Three-dimensional lower bound limit analysis using nonlinear
programming. Ph.D Thesis, University of Newcastle, Australia
References

- 213 -
Lyamin, A. V. and Sloan, S. W. (2002a). Lower bound limit analysis using nonlinear
programming. International J ournal for Numerical Methods in Engineering, 55(5),
573-611
Lyamin, A. V. and Sloan, S. W. (2002b). Upper bound limit analysis using linear finite
elements and non-linear programming. International J ournal for Numerical and
Analytical Methods in Geomechanics, 26, 181-216
Martin, G. R. and Chen, C. Y. (2005). Response of piles due to lateral slope movement.
Computers and Structures, 83(8), 588-598
Matsui, T. and San, K. C. (1992). Finite element slope stability analysis by shear
strength reduction technique. Soils and Foundations, 32(1), 59-70
Miao, L. F., Goh, A. T. C., Wong, K. S., and Teh, C. I. (2006). Three-dimensional finite
element analyses of passive pile behaviour. International J ournal for Numerical
and Analytical Methods in Geomechanics, 30(7), 599-613
Michalowski, R. L. (1989). Three-dimensional analysis of locally loaded slopes.
Geotechnique, 39(1), 27-38
Michalowski, R. L. (1995). Slope stability analysis: a kinematical approach.
Geotechnique, 45(2), 283-293
Morgenstern, N. R. and Price, V. E. (1965). The analysis of stability of general slip
surfaces. Geotechnique, 15(1), 79-93
Morgenstern, N. R. (1992). The evaluation of slope stability A 25 year perspective,
stability and performance of slopes and embankments II. Geotechnical Special
Publication No. 31, ASCE, 1-26
Murthy, B. R. S., Babu, G. L. S., and Srinivas, A. (2002). Analysis of prototype soil-
nailed retaining wall. Ground Improvement, 6(3), 129-136
References

- 214 -
Naylor, D. J . (1982). Finite elements and slope stability. Numerical Methods in
Geomechanics, Proceedings of the NATO Advanced Study Institute, University of
Minho, Braga, Portugal, 1981, edited by Martins J . B., 229-244
Ng, C. W. W. and Shi, Q. (1998). A numerical investigation of the stability of
unsaturated soil slopes subjected to transient seepage. Computers and Geotechnics,
22(1), 1-28
Ng, C. W. W. and Zhang, L. M. (2001). Three-dimensional analysis of performance of
laterally loaded sleeved piles in sloping ground. J ournal of Geotechnical and
Geoenvironmental Engineering, 127(6), 499-509
Ng, C. W. W., Zhang, L. M., and Ho, K. K. S. (2001). Influence of laterally loaded
sleeved piles and pile groups on slope stability. Canadian Geotechnical J ournal,
38(3), 553-566
Pan, J . L., Goh, A. T. C., Wong, K. S., and Selby, A. R. (2002). Three-dimensional
analysis of single pile response to lateral soil movements. International J ournal for
Numerical and Analytical Methods in Geomechanics, 26(8), 747-758
Patra, C. R. and Basudhar, P. K. (2005). Optimum design of nailed soil slopes.
Geotechnical and Geological Engineering, 23, 273-296
Pedley, M. J . (1990). The performance of soil reinforcement in bending and shear.
Ph.D. Thesis, University of Oxford
Pedley, M. J ., J ewell, R. A., and Milligan, G. W. E. (1990a). A large scale experimental
study of soil-reinforced interaction Part I. Ground Engineering, 23(6), 44-50
Pedley, M. J ., J ewell, R. A., and Milligan, G. W. E. (1990b). A large scale experimental
study of soil reinforcement interaction Part II. Ground Engineering, 23(7), 45-49
Plumelle, B. C., Schlosser, F., Delage, P., and Knochenmus, G. (1990). French national
research project on soil nailing Clouterre. Design and performance of earth
retaining structures: proceedings of a conference / sponsored by the Geotechnical
References

- 215 -
Engineering Division of the American Society of Civil Engineers in cooperation
with the Ithaca Section, ASCE, Cornell University, Ithaca, New York, J une 18-21,
1990 ; edited by Philip C. Lambe and Lawrence A. Hansen, 660-675
Poulos, H. G. (1973). Analysis of piles in soil undergoing lateral movement. J ournal of
the Soil Mechanics and Foundation Division, ASCE, 99(SM5), 391-406
Poulos, H. G. and Davis, E. H. (1980). Pile Foundation Analysis and Design. J ohn
Wiley & Sons, New York, etc.
Poulos, H. G. (1995). Design of reinforcing piles to increase slope stability. Canadian
Geotechnical J ournal, 32(5), 808-818
Pradhan, B., Tham, L. G., Yue, Z. Q., J unaideen, S. M., and Lee, C. F. (2006). Soil-nail
pullout interaction on loose fill materials. International J ournal of Geomechanics,
6(4), 238-247
Rahardjo, H., Ong, T. H., Rezaur, R. B., and Leong, E. C. (2007). Factors controlling
instability of homogeneous soil slopes under rainfall. J ournal of Geotechnical and
Geoenvironmental Engineering, 133(12), 1532-1543
Raju, G. V. R., Wong, I. H., and Low, B. K. (1997). Experimental nailed soil
walls. Geotechnical Testing J ournal, 20(1), 90-102
Reese, L. C., Wang, S. T., and Fouse, J . L. (1992). Use of drilled shafts in stabilising a
slope. In stability and performance of slopes and embankments II. Edited by
Seed R. B. and Boulanger R. W. American Society of Civil Engineers, Vol. 2,
1318-1332
Sabhahit, N., Basudhar, P. K., and Madhav, M. R. (1995). A generalized procedure for
the optimum design of nailed soil slopes. International J ournal for Numerical and
Analytical Methods in Geomechanics, 19, 437-452
Schlosser, F. and Guilloux, A. (1981). Le frottement dens les sols. Revue Francaise de
Geotechnique, No.16, 65-77
References

- 216 -
Schlosser, F. (1982). Behaviour and design of soil nailing. Proc. Symp. on Recent
Developments in Ground Improvement Techniques, Bangkok, Thailand, 399-413
Schlosser, F. (1991). The multicriteria theory in soil nailing. Ground Engineering, Vol.
24, November 1991, 30-39
Schlosser, F., Unterreiner, P., Plumelle, C., and Benoit, J . (1992). Failure of a full scale
experimental soil nailed wall by reducing the nails lengths (French research project
Clouterre). Proc. Int. Symp. on Earth Reinforcement Practice, Fukuoka, Kyushu,
J apan, Vol.1, 531-535
Sheahan, T. C. and Ho, C. L. (2003). Simplified trial wedge method for soil nailed wall
analysis. J ournal of Geotechnical and Geoenvironmental Engineering, 129(2), 117-
124
Shen, C. K., Herrmann, L.R., Romstad, K. M., Bang, S., Kim, Y. S., and DeNatale, J . S.
(1981a). In situ earth reinforcement lateral support system. Department of Civil
Engineering, University of California, Davis, Report No. 81-03
Shen, C. K., Bang, S., Romstad, K. M., Kulchin, L., and DeNatale, J . S. (1981b).
Ground movement analysis of an earth support system. J ournal of the Geotechnical
Engineering Division, 107, 1609-1624
Shen, C. K., Bang, S., Romstad, K. M., Kulchin, L., and DeNatale, J . S. (1981c). Field
measurements of an earth support system. J ournal of the Geotechnical Engineering
Division, 107, 1625-1642
Shukha, R. and Baker, R. (2003). Mesh geometry effects on slope stability calculation
by FLAC strength reduction method linear and non-linear failure criteria. 3rd
International Conference on FLAC and Numerical modeling in Geomechanics,
Sudbury, Ontario, Canada, 109-116
Shukha, R. and Baker, R. (2008). Design implications of the vertical pseudo-static
coefficient in slope analysis. Computers and Geotechnics, 35, 86-96
References

- 217 -
Sloan, S. W. (1988). A steepest edge active set algorithm for solving sparse linear
programming problems. International J ournal for Numerical Methods in
Engineering, 26, 2671-2685
Sloan, S. W. (1989). Upper bound limit analysis using finite elements and linear
programming. International J ournal for Numerical and Analytical Methods in
Geomechanics, 13, 263-282
Sloan, S. W. and Kleeman, P. W. (1995). Upper bound limit analysis using
discontinuous velocity fields. Computer Methods in Applied Mechanics and
Engineering, 127, 293-314
SLOPE/W Engineering Book. (2004). Stability modelling with SLOPE/W. GEO-
SLOPE/W International, Ltd.
Smith, I. M. and Su, N. (1997). Three-dimensional FE analysis of a nailed soil wall
curved in plan. International J ournal for Numerical and Analytical Method in
Geomechanics, 21, 583-597
Song, E. (1997). Finite element analysis of safety factor for soil structures. Chinese
J ournal of Geotechnical Engineering, 19(2), 1-7 (in Chinese)
Spencer, E. (1967). A method of analysis of the stability of embankments assuming
parallel inter-slice forces. Geotechnique, 17(1), 11-26
Stocker, M. F., Korber, G. W., Gassler, G., and Gudehus, G. (1979). Soil nailing.
International Conference on Soil Reinforcement, Paris, 469-474
Stocker, M. F. and Riedinger, G. (1990). The bearing behaviour of nailed retaining
structures. Proc. Conf. on Design and Performance of Earth Retaining Structures,
Geotechnical Special Publication No. 25, Ithaca, USA, 612-628
Su, L. J . (2006). Laboratory pull-out testing study on soil nails in compacted completely
decomposed granite fill. PhD thesis, March 2006, The Hong Kong Polytechnic
University
References

- 218 -
Su, L. J ., Chan, T. C. F., Shiu, Y. K., Cheung, T., and Yin, J . H. (2007). Influence of
degree of saturation on soil nail pull-out resistance in compacted completely
decomposed granite fill. Canadian Geotechnical J ournal, 44, 1314-1328
Tabrizi, K. S., Gucunski, N., and Maher, M. H. (1995). 3-D FEM analysis of excavation
of a soil-nail wall. Computing in civil engineering: proc. 2
nd
Congress held in
conjunction with A/E/C Systems '95, Atlanta, Georgia, 812-819
Terzaghi, K. and Peck, R. B. (1967). Soil Mechanics in Engineering Practice. J ohn
Wiley and Sons, Inc., New York
Thompson, S. R. and Miller, I. R. (1990). Design, construction and performance of a
soil nailed wall in Seattle, Washington. Design and Performance of Earth Retaining
Structures, ASCE Geotechnical Special Publication No. 25, 629-643
Tsaparas, I., Rahardjo, H., Toll, D. G., and Leong, E. C. (2002). Controlling parameters
for rainfall-induced landslides. Computers and Geotechnics, 29, 1-27
Tsui, T. Z. (2007). Physical modelling of 3-D and 2-D soil-nailed slopes in Hong Kong.
Final year project report, April 2007, The Hong Kong Polytechnic University
Turnbull, W. J . and Hvorslev, M. J . (1967). Special problems in slope stability. J ournal
of the Soil Mechanics and Foundation Division, ASCE, 93(SM4), 499-528
Ugai, K. (1985). Three-dimensional stability analysis of vertical cohesive slopes. Soils
and Foundations, 25(3), 41-48
Ugai, K. (1988). Three-dimensional slope stability analysis by slice methods.
Proceedings of the Sixth International Conference on Numerical Methods in
Geomechanics, Innsbruck, 11-15 April, edited by G. Swoboda, 1369-1374
Ugai, K. and Leshchinsky, D. (1995). Three-dimensional limit equilibrium and finite
element analysis: a comparison of results. Soils and Foundations, 35(4), 1-7
Unterreiner, P., Benhamida, B., and Schlosser, F. (1997). Finite element modelling of
the construction of a full-scale experimental soil-nailed wall. French National
References

- 219 -
Research Project CLOUTERRE. J ournal of Ground Improvement, 1(1), 1-8
Wang, Y. J . (2001). Stability analysis of slopes and footings considering different
dilation angles of geomaterial. Ph.D. thesis, The Hong Kong Polytechnic
University
Wang, Y. J ., Yin, J . H., and Lee, C. F. (2001). The influence of a non-associated flow
rule on the calculation of the factor of safety of soil slopes. International J ournal
for Numerical and Analytical Methods in Geomechanics, 25, 1351-1359
Wang, M. C., Wu, A. H., and Scheessele, D. J . (1979). Stress and deformation in single
piles due to lateral movement of surrounding soils. Behavior of Deep
Foundations, ASTM 670, Raymond Lunggren, Ed., American Society for Testing
and Materials, 578-591
Weibull, W. (1951). A statistical distribution function of wide applicability. J ournal of
Applied Mechanics, ASME, 18(3), 293-297
Won, J ., You, K., J eong, S., and Kim, S. (2005). Coupled effects in stability analysis of
pile-slope systems. Computers and Geotechnics, 32(4), 304-315
Xie, M., Esaki, T., Zhou, G., and Mitani, Y. (2003). Geographic information systems-
based three-dimensional critical slope stability analysis and landslide hazard
assessment. J ournal of Geotechnical and Geoenvironmental Engineering, ASCE,
129(12), 1109-1118
Xie, M., Esaki, T., and Cai, M. (2004a). A GIS-based method for locating the critical
3D slip surface in a slope. Computers and Geotechnics, 31, 267-277
Xie, M., Esaki, T., and Zhou, G. (2004b). GIS-based probabilistic mapping of landslide
hazard using a three-dimensional deterministic model. Natural Hazards, 33, 265-
282
References

- 220 -
Xie, M., Esaki, T., and Cai, M. (2006a). GIS-based implementation of three-
dimensional limit equilibrium approach of slope stability. J ournal of Geotechnical
and Geoenvironmental Engineering, ASCE, 132(5), 656-660
Xie, M., Esaki, T., Qiu, C., and Wang, C. (2006b). Geographic information system-
based computational implementation and application of spatial three-dimensional
slope stability analysis. Computers and Geotechnics, 33, 260-274
Yamagami, T., J iang, J . C., and Ueno, K. (2000). A limit equilibrium stability analysis
of slopes with stabilizing piles. Slope stability 2000: proceeding of sessions of
Geo-Denver 2000, August 5-8, 2000, Denver, Colorado, Sponsored by the Geo-
Institute of the ASCE, edited by D. V. Griffiths, Gordon A. Fenton, Timothy R.
Martin, 343-354
Yang, M. Z. and Drumm, E. C. (2000). Numerical analysis of the load transfer and
deformation in a soil nailed slope. Numerical methods in geotechnical engineering
recent developments, Proceedings of sessions of geo-denver 2000, Geotechnical
special publication No. 96, August 2000, Denver, Colorado, 102-115
Yuan, J . X., Yang, Y. W., Tham, L. G., Lee, P. K. K., and Tsui, Y. (2003). New
approach to limit equilibrium and reliability analysis of soil nailed walls.
International J ournal of Geomechanics, 3(2), 145-151
Zhang, M. J ., Song, E. X., and Chen, Z. Y. (1999). Ground movement analysis of soil
nailing construction by three-dimensional (3-D) finite element modeling (FEM).
Computers and Geotechnics, 25(4), 191-204
Zhang, L. M., Ng, C. W. W., and Lee, C. J . (2004). Effects of slope and sleeving on the
behavior of laterally loaded piles. Soils and Foundations, 44(4), 99-108
Zhang, X. (1988). Three-dimensional stability analysis of concave slopes in plan view.
J ournal of Geotechnical Engineering, 114(6), 658-671
References

- 221 -
Zhang, X. (1999). Slope stability analysis based on the rigid finite element method.
Geotechnique, 49(5), 585-593
Zheng, H. (2007). A rigorous three-dimensional limit equilibrium method. Chinese
J ournal of Rock Mechanics and Engineering, 26(8), 1529-1537 (in Chinese)
Zheng, Y. R., Zhao, S. Y., Kong, W. X., and Deng, C. J . (2005). Geotechnical
engineering limit analysis using finite element method. Rock and Soil Mechanics,
26(1), 163-168 (in Chinese)
Zhu, D. Y. and Qian, Q. H. (2007). Rigorous and quasi-rigorous limit equilibrium
solutions of 3D slope stability and application to engineering. Chinese J ournal of
Rock Mechanics and Engineering, 26(8), 1513-1528 (in Chinese)
Zhu, D. Y. and Lee, C. F. (2002). Explicit limit equilibrium solution for slope stability.
International J ournal for Numerical and Analytical Methods in Geomechanics,
26(15), 1573-1590
Zhu, D. Y., Lee, C. F., and J iang, H. D. (2003). Generalized framework of limit
equilibrium methods and numerical procedure for slope stability analysis.
Geotechnique, 53(4), 377-395
Zienkiewicz, O. C., Humpheson, C., and Lewis, R. W. (1975). Associated and non-
associated visco-plasticity and plasticity in soil mechanics. Geotechnique, 25(4),
671-689

You might also like